It should be posted here, eventually:
http://www.ws100.com/medicalresearch.htm

Thanks Mark

Sorry for the multiple posts. My brother (who is better at math) calculates 9.8%, 19.5%, 29.2%.

Without getting into too much detail... Even with 3 tickets, the odds are not significantly better, overall. The pool of 3-time applicant tickets (774) in the "bucket" is still substantially less than the number of 1-time applicant tickets (1,224), and still less than the 2-time appplicant tickets (922). The key is there are overwelmingly more 1 & 2- time applicant tickets. Net: there will be more 1st time ticket holders selected than 3rd time ticket holders. The only thing that improves statistically, which is nil, pertains to ticket holders within each of the 3 categories.There will higher percentage of 3-time ticket holders selected within it's own category (not overall combined categories), but that is due to the small number of 3-time tickets compared to 1-time tickets. In the end, it's purely luck of the draw, statistically speaking. <br />-signed, <br />4- time consecutive (sans fire year) entrant

Larry - I completely agree with your analysis. The three-ticket holders, as a group, have the worst chances of being selected, due exactly to what you mention, the preponderance of one-ticket and two-ticket holders. Some might be interested in the results of two mock lotteries based on tallies earlier in this year's lottery signup. One mock result showed that while only 10 percent of the one-tickets holders were selected, over 42 percent of the selected entrants were one-ticket holders, and that while 27 percent of the three-ticket holders were selected, only 24 percent of the selected entrants were three-ticket holders. Here are the results:
https://sites.google.com/site/twentyfivesquared/mock-western-states-lottery-shows-that-three-ticket-holders-have-the-worst-chances

I agree with the premise -- that it is unfair for some runners by chance to fail to get in for multiple years, whereas another runner barely qualified to run 50 miles gets in first time. However, the article referenced above confuses distribution with probability. The article referenced above states that 6.35% of first-timers will get in and 3.62% of third timers will get in, or odds of 1 in 30. This fails a basic test of logic. If that were true, I should talk the lottery into removing two of my three tickets, thus increasing my odds. The fact is that a first-timer has odds of 1 in 10, and a third-timer has odds of 1 in 4.

so if there was a theoretical 4th group which had only 10 people in it, but each of them had a 100% certain chance of getting in, they would as a group only get 3+% of the tickets ....so with the same logic, would this group (as a group) have even lower chances ?

I agree, David, that the results of the mock lottery defy common sense. In one of the two mock lotteries, the final breakout gave three-ticket holders a 3.62% chance of being selected (73 in 2017), the two-ticket holders a 4.91% percent chance (99 in 2017), and the one-ticket holders a 6.35% chance (128 in 2017), with overall chances of 14.87% (the sum of the chances). Those results are based on the person count at the time and an assumption of 300 entrants.

Should anyone really care about the number of people from your group that get in or what percentage of the total that is? This is not the House of Representatives or something where the biggest group selected has an advantage...it's irrelevant here. Of course the group that has by far the most people will still have the most members selected even if they have individually lower odds. All that really matters here is individual odds...and based on my math, having more tickets in the hat as an individual increases your percentages to approximately what David and his brother calculated. Good luck to you all tomorrow!

Group odds are not the most important thing, that's for sure, but they might get us thinking about what kind of advantage the multi-ticket holders really have. I think very little advantage, and certainly not worth all the hype. The main point about the two mock lotteries is that the results do not support that three-ticket holders have three times the chances of being selected as one-ticket holders. For that to be true, the following would need to be true (using round numbers): 300 people will be selected from 3000 tickets, or 10%, so people who have two tickets will have a 20% chance of getting in, and people with three tickets a 30% chance. That's a false leap. The only level at which a three-ticket holder has three times the odds of a one-ticket holder is the level of the individual draw, where the odds, on the first draw, are indeed dismal - 1 in 3000 or 2 in 3000 or 3 in 3000, depending on you ticket count - and hardly better on the final draw. To put this another way, overall the odds are about 1 in 10, but to conclude from this that the odds are 30% for the three-ticket holders and 20% for the two-tickets holders would mean that three times as many three-ticket holders get selected as one-ticket holders, which could not happen in this system by random selection. The important thing, in my opinion, is that multiple tickets do practically nothing to increase your chances of getting selected, unless you have a huge number of tickets, which is what we sometimes see with the raffles.

" ...would mean that three times as many three-ticket holders get selected as one-ticket holders, which could not happen in this system by random selection."
There's still a big logic problem here. As per John's example, if you had a fourth group of ten applicants with four tickets each, you could never have four times as many of the fourth group selected as the first group selected.
If you must use the mock lottery to figure the odds, run it a hundred times at least and see how many times an individual ticket gets selected.

Brian, absolutely everyone should care. It just doesn't make any logical sense that there will be more 1 ticket people drawn than 3 ticket people. Not sure how anyone can debate that, but you can sure try.

Just to be clear - you guys are saying that an individual person with 3 tickets has LESS of a chance than an individual person with 1 ticket? That is your conclusion? If so, feel free to give me your extra tickets so you can have 1 ticket and a better chance.. some people are saying very little advantage, but the post and the 'mock' lottery is saying LESS chance for a specific individual? Huh. To the last post, it would make more sense that more 1 ticket people drawn if there were more total tickets of 1 ticket people than total tickets of 3 ticket people.

Excellent point, Joe. I am going to keep my extra tickets just in case you are right. I believe there are multiple ways to determine these odds, but here's my logic and math:

First, the facts:
• One-time lottery entrants: 1,222
• Two-time lottery entrants: 460 (two-time entrants get two names in the hat)
• Three-time lottery entrants: 258 (three-time entrants get three names in the hat)
• Total lottery entrants: 1,939
• Total names in the lottery hat: 2,916 (1,222 + 460*2 + 258*3 = 2,916)
• Number of people selected in lottery: 265 spots

The chances of a one-time lottery entrant getting selected are pretty easy, it’s just 265/2,916, which is 9.09%.

Now for those with two names, we start by figuring the odds that you are not selected with your first chance. With 2,916 names in the hat and 265 pulls, that equals 2,651/2,916 – or 90.9122%. Now the odds of not being selected with your second name are a very tiny bit more favorable at 2650/2,916 – or 90.8779%. Now to find the odds that one of your two names is selected, we multiply both and subtract from 1. So, since .909122 * .908779 = .8261912, after you subtract that from 1, that means all of the two-time entrants have a 17.38% chance of getting selected.

Now for those with three names in the hat, just multiply the odds of not getting drawn after two pulls (.8261912) by the odds of not getting drawn with your 3rd chance (2,649/2,916 = .908436) and you get .750542. Subtract that from 1 and your odds of getting drawn with three names in the hat are 24.9458%. Approximately.

Disclaimer: This doesn’t factor in that a person can only be drawn once, and once someone with multiple tickets is drawn, their additional tickets are essentially removed from the hat. This fact would slightly increase these odds, but not by a lot.

Lets complicate the odds even more cuz this is so much fun - what about the fact the last ten or so tickets are drawn from people who are present at the lottery - what are their odds? What if you are a three ticket holder and present vs a one ticket holder, (and now we must switch the above odds to take into account there are really only 255 spots for most applicants!!)
Just stirring the pot a little.

That is not what we are saying. What we are saying is that the odds are that there will be more 1 ticket holders chosen than 3 ticket holders. I understand that individually my odds are 3 times better against a single person with one ticket. I will keep my extra 2 tickets....thank you :)

Probabilities sure are tricky. The 9.09% figure that Brian gives is the sum of 265 1/2916 probabilities, or 265 * (1/2916). So we would say that the 9.09% figure represents overall odds for people with one ticket, not odds on any one draw, which are vanishingly small. I would bet a year's supply of running shoes that around 40% of the selected runners will one-ticket holders with the remaining roughly split between the multi-ticket holds and three-ticket holders getting the fewest spots of all. I stand fast that the current lottery system does not meaningfully improve odds for multi-ticket holders. If anyone is interested in tracking the real outcome, I would suggest grabbing a copy of the current rosters today then matching names after the fact, because I would also bet that after the lottery we will lose visibility on who had how many tickets. This goes to the transparency theme, much more important, in my opinion, than the math. I don't care so much what the system is as long as it is known. One step in the right direction would be to publish the selection algorithm for public comment.

i think it is a binary event ... you are either going to get in OR you aren't .... good luck to all !!

My predictions: 114 1st timers, 82 2nd timers and 66 3rd timers will be picked, using the odds I've calculated. So yes, Thomas, over 40% will be 1st time ticket holders. Still, 3rd timers have a one-in-four chance, considerably better than first timers. 258 / 4 = 65. If the numbers are wildly different, my math is off or the system is not as transparent as I think it is.

David - It looks like we are in close agreement. Your breakout closely matches the results of the mock lotteries. I think that we should clarify that while these numbers give the three-ticket holders a 1-in-4 chance both within their group and overall among the selected, this group still has the worst results in terms of percent of total entered (1941) and total selected (262). I guess it would just feel better to me if this group had the best results. I think that if we are going to bias a system, we should bias it meaningfully. For example, let's clear half the three-plus-ticket list every year. If we need a separate lottery for that group, let's try it. Let's try something. While we are at it, let's enhance the transparency of the process by publishing, for each selected entrant, the means of entry. Extending your numbers:
Selected % of Selected % of Entered
114 43.51% 5.87%
82 31.30% 4.22%
66 25.19% 3.40%
262

David is right on with the estimates for the number in each group which will be selected. Although it doesn't appear like it, 3 ticket holders have approximately 3x(slightly lower) the probability of getting selected compared to 1 ticket holders. You may not like it that there are more first timers selected than 3x people, but that has nothing to do with the fact you have significantly better odds with 3 tickets. It is possible that the board should consider a greater weighting to multi year losers. Such as a 2^n progression 1,2,4,8,16. This will help get an even higher percentage for multiple time losers. The major issue is that supply far outweighs demand. This year there are 1944 entrants and only 265 lottery slots. It would take 7.3 years to clear out all the names at this rate. The board can also consider making the qual standard tougher to reduce the number of people in the lottery. I'm keeping my 3 tickets and hoping for good luck tomorrow.

You guys are all nuts. Really. Keep your fingers crossed, that's all you can do, 3 tickets or 1. Another option is to get through different accepted venue. Tomorrow. One more day for y'all.

The Western States 2012 lottery is complete. Of the 1940 people entered in the lottery, 267 were selected prior to the few spots reserved for a drawing among people who attended the lottery.

I can already hear the calls to "lighten up" on this topic, but signs that the lottery is not conducted by the fairest means possible continue to bother me.

Today's results are so skewed against what would be expected in a random selection as to suggest either a bug in the selection algorithm, lack of sophistication in the algorithm, or even manipulation of the lottery.

Of the 267 people selected, 154 had one ticket in the lottery, 71 had two tickets, and 42 had three tickets. This means that 57.68% of the selected had one of the 2926 tickets, 26.59% had two of the 2926 tickets, and 15.73% of the selected had three of the 2926 tickets. In comparison: Among the 1940 applicants, 41.88% had one ticket, 31.60% percent had two tickets, and 26.53% had three tickets. Statistically, this is a very large discrepancy and suggests a selection process that is not adequately or fairly random (to put it objectively).

The odds within each group broke out like this: 12.60% of the one-ticket holders were selected, 15.43% of the two-ticket holders, and 16.28% percent of the three-ticket holders.

Clearly, having multiple tickets does not represent a significant advantage. The nearly 84% of the three-ticket holders who were not selected today should be upset by this inadequate and unfair system.

In the spirit of transparency, since it is unlikely that we will see a breakout of selected by ticket counts on Ultra Signup or anywhere else, I will then explain how I came up with these numbers. I would also like to mention that was not entered in the lottery.

1) At the close of the selection process, I copied the table of names from the ultralive.net feed into an Excel spreadsheet where I already had a list of all applicants and their ticket counts.

2) I used an Excel function called VLOOKUP to match the name of an applicant with a name of an entrant and return the number of tickets that the applicant held in the lottery. I performed a sanity check on the names, making sure that every exact name in the list of selected exactly matched a name in the list of applicants.

3) Among the selected, I counted the number of people who started with one ticket, two tickets, or three tickets. After that, I could do the percentages.

I am also quite sure of at least one case of someone who was chosen in last year’s lottery and, with two tickets in this year’s lottery, chosen again. How would that be possible? Not that having extra tickets makes that much difference, after all. But If race leadership would plainly and transparently publish means of entry, the public would have a chance to catch such problems that, without evidence otherwise, appear as favoritism.

Thomas, I have tried to make this as transparent as possible. The list of applicants has been made available for everyone to see and catch anything that looks suspicious. Contrary to your inference, there's not a team of people working on this, This is ultra running, not Ironman. I welcome all suggestions and help identifying issues. I have absolutely zero time, desire or motivation to manipulate the data. Seems you have some extra time, would you care to create an algorithm we could test out? Feel free to contact me at mark@ultrasignup.com and we can test it out for next year's event.

Mark - Thank you for your reply. I would be happy to help. I can imagine how busy you must be maintaining Ultra Signup on top of everything else. Perhaps we would be able to pull together a team on this topic to share the load. I have some ideas. There are multiple Open Source code projects as well as some paid services that companies use to hold impartial lotteries. I will contact you. Tom

If you would like to review the current methodology, you can Google: SQL Server, NEWID(), order by. This is a wildly accepted method of selecting random records from a database. This and almost all methods have extensive discussions out there.
Any solution you would like to create and test will have to be database driven aka SQL.

I will check it out! Thank you for making this public (not that you were resisting that) for anyone to check out. Yes, SQL assumed.

Some might be interested in how the New York Road Runners hold lotteries for their very popular half and full marathons, for which the chances of being selected by non-guaranteed entry are probably even worse than the chances in the Western States lottery. The NYRR use a paid service called random.org to generate their picks. The point is that the wheels have already been invented and we just need to test them out. I know that the process would not be that simple, but I think we should try.

Mark - One thing I did not realize is that the WS Board has given Ultra Signup so much control in this area. Like you said, however, this is ultra-running. Just the same, as the popularity grows, the rigor of entrant selection should also grow. What I would really like to see on Ultra Signup is a new column on the WS Entrants page listing means of entry (lottery, in-person lottery drawing, sponsor, aid station, Montrail Cup, going for tenth, top ten, and whatever other categories there might be).

http://www.nyrr.org/races/2012/nychalf/non_guaranteed_entry.asp
http://www.random.org/

First of all, I accept the premise of "their race, their rules", however, a couple of observations if I may. I am one of about 200 three-time-losers facing about a 75% chance of becoming a four-time-loser this time next year. It seems apparent that unless a system is devised which somehow 'clears the board' of the biggest losers, then it is almost certain that in seven years time, there will be some people who will have lost 10 times in a row, looking forward to having 11 tickets in the draw for 2019's lottery .

My only suggestions are - instead of adding one ticket each time you lose - what about doubling up? Or, perhaps, next year making (for example) 40% of the spots available to 3-time losers, 30% to 2-time losers, 20% to 1-time losers and first time applicants get 10% of the race spots.

I appreciate your efforts Mark, you have a great website!

This is probably the wrong forum to discuss many of these ideas as Mark can only follow the rules he is given by Western States. Unfortunately because he runs the actual registration he is on the receiving end of the complaints. Ultra Signup does a great job and I hate to see them being blamed for a rather unwieldy lottery. Mark, do you know if the WS Board follows this forum at all? The old three strikes and your in policy was very popular and I think it would have been better to tweak that then set up this new system that really doesn't give repeat applicants a significant advantage. How about giving the highest group of ticket applicants half the draw numbers, and then just have a normal draw from the remaining? It's especially hard for the group at the upper end of the age groups as the number of years we can keep doing this is limited.

Indeed, Mark is doing a GREAT job. Now it's time for me to eat some crow. I have been working with Mark this morning on the lottery algorithm (see posts above). The algorithm he uses is well established for providing a random result, which got me thinking that my numbers above, concerning a statistically high proportion of selected one-ticket runners, might be wrong. I reviewed my own work and I found a mistake. The bottom line is that the numbers from yesterday's lottery closely match the results of the mock lotteries discussed in this thread. Yesterday, 43.45% of the selected runners had one ticket, 34.83% had two tickets, and 21.72% had three tickets. Compare this to the result of the mock lottery (link above): 43,67, 33.00, and 24.83. That is close. Whether or not the result is fair to the multi-ticket holders I think is a matter for a different discussion. Mainly I want to say that yesterday's results show no signs of manipulation but rather closely reflect what we would predict from a random selection. I think it is GREAT for what we do that the algorithm is now published and available for comment. That is transparency! I will echo my call for the means of entrant selection to be published also (lottery, raffle, Montrail, top ten, going for tenth, board member, aid station, sponsor, special consideration, and whatever others).

Sorry, everyone, for having ruffled any feathers in my zeal for transparency in the process. Just the same, we really have made some progress on this topic and now we can work on whether the multi-ticket approach is worth pursuing. We can work together to get the attention of the WS board. We have more influence that perhaps we realize. The results - consistent now between the real and mock lotteries - do demonstrate that we are not clearing the decks of the unlucky unselected. I think that Mike's "power of 2" suggestion above is definitely worth further review. This would mean, instead of one more ticket for each consecutive year not selected, you would get a power-of-two number of tickets, which is to say 2, 4, 8, 16, 32, etc., instead of 1, 2, 3, 4, 5. I also like James's idea of separate lotteries for each group of runners.

"The important thing, in my opinion, is that multiple tickets do practically nothing to increase your chances of getting selected, "

I picked this as a summary quote from some of the arguments above, with some people actually unsure if having extra tickets was helpful, but it's absolutely false. The worst outcome of this discussion would be for people's understanding of math and probability to be further distorted. Having 2 or 3 tickets absolutely, unequivocally, and inarguably doubles or triples (respectively) one's individual chances of getting picked. It's refreshing that a single mock lottery approximated established probabilistic theory, even though multiple simulations (as suggested above) would have been required to increase the significance of the results, but the odds were easily calculated, as shown in previous calculations above.

To better state the debate succinctly and accurately, the perceived unfairness -- and I agree with the sentiment -- is the fact that the overall baseline odds are increasingly dominated by the shear number of new entrants, and many of has have a perception that entering 3,4,5+ times is very hard work, in a limited window of opportunity in our lifetime. I get that, especially because we might perceive a person who qualifies 3-4 years consecutively as having been proven to work consistently at the goal. At the same time, if we slant heavily toward multiple entrants, multiple consecutive qualifications will become the new de facto entrance requirement.

With the current number of applicants, it is still probable that somebody entering 4 times in a row will get selected for the lottery, precisely because of the reasonable odds of having gotten picked the first, second, third, or fourth year. (Very roughly, a person would have a 0.9*0.8*0.7 = 50% chance of not getting picked 3 years in a row). The crux is, once somebody gets to less likely 3-or-4 ticket state, his individual odds improves year-over-year at a rate less than before: going from 1 ticket to 2 effectively doubles one's chances, whereas going from 4 to 5 only increases the year-over-year odds by 25%. In that respect, I see the arguments for exponential and/or group allocation.

It is a delicate balance of worthy of discussion, but I wanted to make sure the discussion of probability theory was at least accurate. If people are misunderstanding the current lottery algorithm, any future changes will be subject to criticism as well!

I agree, Michael, with almost everything you said. You said the chances of getting picked in three lotteries is about 50%. However, if you have not been picked twice in a row then the chances of getting picked with three tickets is still slim as those odds are not dependent upon the previous two lotteries.
My proposal would be to tighten up the entrance requirements, as they are trying to do now. To significantly tighten them would eliminate some mid-to-back of the pack runners who may be able to finish WS in 30 hours. So instead, why not require another hundred as qualification? I'm not sure, as constrained as the race is, that it is right for someone to enter this as a first hundred.

I certainly agree that the probability of not getting selected three years in a row is the product of the probabilities of not getting selected in each of the three years. The real lottery results show this, for the 267 selected before the had-to-be-in-the-audience raffle and the "special consideration" entries:

One-ticket holders: 116 selected or 43.45 percent
Two-ticket holders: 93 selected or 34.83 percent
Three-ticket holders: 58 selected or 21.72 percent

This means that the one-ticket holders had, overall, a 56.55 percent chance of not getting selected, the two-ticket holders a 65.17 percent chance, and the three-ticket holders a 78.28 percent chance of not getting selected.

I would like to pause here and suggest that something is wrong with that progression: the more tickets you have, the higher your chances of not getting selected. I think it is misleading to look at percent within group as opposed to percent of total runners.

Now, if you take the product of 56.55 percent, 65.17 percent, and 78.28 percent, the result is 28.85 percent. This means that the overall chances of not getting selected three years in a row, based on the one set of numbers we have, is 28.85 percent, or a 71.15 percent of getting selected once in three years. If we just give that percentage, and ignore the progression of chances in the wrong direction from the point of view of the multi-ticket holders, I think that we are missing the point. My premise is that people's chances should improve, overall, as they get more tickets.

There is one perspective from which a person's chances do double with two tickets and triple with three: the chances on the first draw. In the last lottery, the chances on the first draw were 1 in 2918 tickets, or 0.0003427 (not to put too fine a point on it). This means, on the first draw, if you have one ticket, your chances are about 34 in 100,000; if you have two tickets, your chances are twice that, or about 68 in 100,000; and if you have three tickets, your chances are three times that, or about 103 in 100,000. Those are really bad odds! To get people thinking and talking that three-ticket holders have three times better chances than one-ticket holders is really to mislead people and to create a lot of confusion in the process. (To determine chances on successive draws gets a little more complicated because we need to account for pulling one, two, or three tickets from the pool, depending on who is selected.)

The whole point of the mock lottery was to dramatize how misleading such thinking can be. I maintain that the bias for multi-ticket holders is not nearly strong enough to create a more level running surface. Of course, the Western States trail is not a level running surface, and, without joking too much, maybe that is part of the point. I wholly concur with the idea of making it more difficult to qualify to apply. The list of qualifying races

I agree (and was trying to keep it separate from the stats analysis) that requiring a 100M qualifier is a good and necessary first step. I voluntarily did not enter the WS lottery after my first couple of 50's, thinking that WS being my first 100 wouldn't be fair to anyone.

Thomas, I again disagree with the third paragraph, above, and the conclusion: "the more tickets you have, the higher your chances of not getting selected." This is false and was the point of my post: more tickets absolutely increases the chance of getting selected. In your percentages, I'm afraid there's some "denominator switching" going on.

116/1222 single-ticket holders were drawn. The reason it was 43% of the overall field is because it was 1222/1939 = 63% of the overall applicants. Single-ticket holders are underrepresented.

58/258=22% of the 3-ticket holders were drawn, despite the fact that 3-time ticket holders were 58/1939 = 3% of the overall applicants. 3-ticket holders are overrepresented, by design.

The feeling is that 3,4, and 5+-ticket holders should have a "very good, almost guaranteed" chance of getting in. This is worthy of debate. But it is absolutely true that multiple ticket holders most definitely have a better chance of getting in!

Why does this dead horse continue to be beaten? Michael, I would argue your opinion is biased since you are on the start list for next June and it's much easier to expouse the great odds under those circumstances. Congrats, by the way! I had 3 tickets, plus I had entered when the TTL rule was still in effect (first lottery was Dec '07)--that's 4 lotteries + the no-lottery after the fire year. I"ve pace from FH to the finish and was at the race in 2010 and watched Geoff and Tony finish within minutes of each other. Over the years, I had become a student of the race. Personally, I'm done entering my name in the lottery. In my mind, I didn't feel validated as an ultrarunner without having that buckle (preferably, the silver one). Thankfully, in my household, we have three buckles. :o) But, life is too short to get so wrapped up in one race every year and I'm moving on.

Michael - I truly applaud the nobility of your patience in waiting to enter the lottery process until you felt you had run an event that qualified you to tackle the challenge. And without question, I agree, multiple tickets for multiple "losers" tempers the dominance of the one-ticket holders. But not enough. What I'm really trying to convey is that the bias is insufficient. The process should support aggressively clearing the lists of the most unlucky. I believe - with all respect to Mark Gilligan, who has put in countless hours to maintain this site - that the Western States Board hastily instituted a half-baked lottery system when it became obvious that soon demand for entry would increase to the point that only two-time losers would get in and, eventually, not even all of them. I don't think that anyone really explored the math of it and really looked into what kind of statistical bias is appropriate.

Larry - to address your question about a dead horse. I stay with this subject because I want to change the system and I want to retain my right to complain about it. I stay with it for the "lottery unlucky" who cannot so easily let go of the drive to participate in the Western States Endurance Run. I stay with it because I cannot tolerate an opaque process, the lack of transparency of the "special consideration" runners. I stay with it because it is simply not right when people who are new to the sport and who ran the weakest of qualifiers magically appear on the entrant list post lottery. If I could achieve one thing (and back to dead horses), I would want to add a simple column to the list of entrants that publicly proclaimed the means of entry. No organization in the public trust should be allowed to operate behind a closed door. I stay with it because I remember the one and only time I made the lottery, what euphoria, what excitement, and now, having done it, what amazement that to this day my aging mind retains astonishingly strong memories of every single hour - I could even say every mile - of the nearly 30 hours that I spent on the trail during those two sunrises. More people should have that experience. The same people should not chronically repeat the experience.

And about the buckle: the least important thing to me in the entire experience. The "buckle" is the indelible mark on my soul of having reached the finish line.

I will add a proposal for one new rule: if you DNF, you cannot enter the lottery for the next year. You might qualify by other means, but you cannot enter the lottery. You had a shot. You didn't make it, for whatever conspiracy of reasons. Tough luck.

Thomas: I wholeheartedly agree with paragraphs 2 & 4. I don't "easily let go" of this whole thing because I have a lot of emotional time invested in it. Just ask my wife (Olga). WS was the first ultra I ever knew existed and was determined to stand at the base of Squaw one day in the dark--bib pinned to my shorts with several hundred runners during the last weekend in June. But, eventually, as the saying goes, you either have to fish or cut bait, and I don't do that easily. At this point I feel like I'm doing nothing but whining about it. Nonetheless, your efforts in trying to improve and influence the system are greatly appreciated.

Gents, Just to weigh in on part of this OTHER than the numbers... I have worked the race as a volunteer the last two years, all the way through from start to take down, completed my qualifying events at the 50 mile level, entered the lottery twice,ATTENDED the lottery twice, and completed the Marathon de Sables this year (156 miles in 120 degree heat) but WS would have been my "first 100".. your opinion then is that I shouldn't be able to be considered??? I am NOT in this year (unless I get selected as a aid station sponsored runner) but I will again run my qualifier, be back there in June volunteering and will be there next December for a third lottery. I think I have earned my right to toe the line at Squaw when my name gets pulled, regardless of where I am in the pack but thats just my .02 ;-)

I'd say it's done Ed,I know they've sent out emails for the special considerations already so I'm presuming the aid station recipients are done also

The captains are being asked to submit their designated runners now. They are usually added through the first part of January. Good luck.

Thanks gents

Sometime this week.

6 out of 115 - only a 5% showing - I agree a few more spots would be kind - but also wonder if foreign applicants should have an advantage over North Americans? Should Californians have a better chance to run in their own race? Lots more subjects for discussion - but at this point it's all conjecture since i won't be running WS this year :(

I have no problem with it being a straight draw but they claim to want global representation.6 runners in a field of over 400 can't really class itself as a global race

Perseverance is the only attribute that gets rewarded. The Board decided to remove all other considerations when they revamped the lottery rules (no more TTLs no more international spots, etc.) As such it is a level playing field for all interested applicants regardless of sex, age, location: a qualifying time gets you in the lottery; the only tweak they chose to make was to give multiple-time applicants a higher chance of selection.
Personally, I don't expect preferential treatment in races overseas, not sure why this should be any different)

All valid comments above, however International athletes are disadvantaged as we can't easily get to the Montrail Cup races and try for a spot runnings 1st or 2nd in a race.

First, international folks must have an address outside of North America. Second, we will analyze their running history to make sure they are not simply living here and using an alternate address. Our database is good at cross referencing past registration addresses. Also, if you know of someone who is trying to trick the system, please let us know. We expect Ultra runners to live up to a higher standard.

May be not the registration address as I can give you one as well, but if a person runs every month in US (say, PNW, or CA, or whatnot, not once a year), wouldn't it be obvious such person lives here for at least most of the time? Just click and go through each name before the drawing so it's more transparent and fare. Also, in line, was wondering if there will be any data digesting of each name who got in to WS - the path to glory so to speak, as in whether they got through lottery (1 ticket or 3?), automatic, international, AS help, special consideration...you know, since it's such a huge deal for all, I am certain people would like to know:)

Everyone in the lottery is on the same playing field (except extra ticket runners). International runners will not have a greater chance of selection. However, if there is a post lottery special consideration, the list of internationals may be considered. I will ensure the board has enough info to identify anomolies and potential cheats. After the selection process, I will TRY to produce a breakdown of the selected folks.

Aw, thanks! I guess it was just for showing who is from where, whether Utah or another country. Sorry. And thanks for hoping to try. And I am just glad I chose not to be one waiting for the lottery Gods this weekend:)

Most likely, the result of flushing through the "old" 2-time losers that they spread across 2 consecutive years. Now that the TTL's ran in 2010, the new ticket system is in effect resulting in a greater number of slots for the lottery applicants, sans automatics.

I believe Doone Watson embodies what "Ultrarunners" are all about ! no matter which way this Lottery goes, personally it's a positive all the way around! Even though i have never been picked in 2003, 2009, 2010 and now 2011. My WS100 Journey started in 1998 after i ran a Sub3 Marathon in the Old Sacramento race and then ran the Boston. I asked myself " what do i do for an encore ?, my question was answered a few weeks later when Greg Soderlund sent me an invitation through the mail to qualify and run WS100. Which meant running 4 times the distance of Boston!
If you accept the challenge mentally, then you will go "ANY" distance!
In a nutshell Greg planted the seed for me to continue running and being healthy, and an even better outcome, as my "Running Resume" grows, i hope to inspire more people to stay active, not only when your young, but to continue until your name is called, by the other Lottery!

As this cold weather settles into Calgary I am thinking I am in a win win situation anyhow - if I get picked I win the lottery, if I don't I win an easy winter and don't have to run in the cold for hours and hours!!!

Michael, how would a "split lottery" work?

Michael is completely making too much sense :) I have been saying this also. Split lottery would work like this. 90 spots to the 1 ticket people. 100 spots to the 2 ticket people. 110 spots to the 3 ticket people. 3 separate lotteries. For 3 ticket people the odds would be approximately be 42%, 2 ticket 21%, 1 ticket 7%. Seems simple and fair as can be.

I agree. Right now it is possible for someone to eventually have ten tickets in the hat and still not get picked. That doesn't seem fair. With a split lottery, you'd eventually get to the point, say with 5 ticket people, where there would be more spots than available entrants. Thus you'd have a de-facto five time loser rule.

if you have 3000 tickets for 300 spots, and the 3 timers are 250 entrants with 750 of the tickets, and the 2 timers are 500 people with 1000 of the tickets, and the first timers are 1250 with 1250, then by equal chance the 3 timers should get 75 spots, the 2 timers should get 100 spots and the first timers should get 125 spots. You have a split lottery in effect.... the extra 2 chances ARE indeed worth a lot. As Doone pointed out below not quite 3 times as good as having a single ticket but around 2.7 times as good... Good luck to all !

True John. Just my opinion that the 3 ticket holders should be rewarded with more spots.

Jeffrey - since i am in 3 times i can't argue with that :)

So to take it further .... that means about 175 of the 3 timers won't get picked this year... if you assume that 150 of them requalify and enter again next year and each have 4 chances ... then maybe about 60 of those get picked, etc. etc. might look something like this ( after 5 years everyone is going to requalify and try again! )

year 3 = 250 enter 75 get picked
year 4 = 150 enter 60 get picked
year 5 = 75 enter 35 get picked (takes 40 spots to include all)
year 6 = 40 enter 20 get picked (takes 20 spots to include all)
year 7 = 20 enter 11 get picked (takes 9 spots to include all)
year 8 = 9 enter 5 get picked (takes 4 spots to include all)
year 9 = 4 enter 2 or 3 picked (takes 1 or 2 spots to include all)
year 10 = 1 or 2 pour souls have a 65% chance to get in

Obviously they should at least put in some sort of "x-year loser" rule on top of the current system.

No

Hello Jason, unfortunately crewing and pacing cannot be used to fulfill the WS service requirement.

Thanks for the info. I am currently working in Afghanistan and if selected I will take vacation to come out for the race. Any suggestions that would fulfill the service requirement? Could I RD my own race out here or volunteer at the USO?

Given your circumstances, yes, you could RD a 5K or 10K fun run to fulfill your service requirement.

This is an excellent website if you are really want to know your odds http://www.wikihow.com/Figure-Your-Odds-of-Holding-a-Winning-Lottery-Ticket
Does take a calculator - remember that if you have one ticket and a 1/10 chance, that doesn't mean having 10 tickets gives you a 10/10 chance!! As an example if there were 3000 applicants, 300 spots available and you had one ticket you would have a 10% chance. If you had three tickets it would increase to 27% chance and if you had 10 tickets it would only increase to 65% chance. Hopefully no one has to apply over ten years before they get in!

It would be a little bit better than this - probably around a 11% chance for one ticket ... because there are only 2000 unique applicants. If a runner with multiple chances is picked, it would remove all of that runners extra chances as well.

Hi Doone, thanks for the interesting analysis. The only thing I like more than running is working with probabilities. The 65% chance of winning with 10 tickets is an interesting result. Some people would guess it has to be closer to 100% but it is far from it. A somewhat similar situation also gives odds of around 65%. Say there is a lottery such as 6/49 where there are 14 million unique combinations. If one were to randomly buy 14 million tickets, then the odds of hitting the jackpot is no where near 100%. It is also around 65%.

This is only meant to show the interesting parallel between the two situations and not prove the original results. The two "games" are very different since the WS100 example is a raffle with guaranteed winners and the second example is a lottery where having any winners is not guaranteed.

Since you have 3 tickets, I hope your name gets pulled early thereby increasing my chances of winning.

Good luck,
Mario

Would be interesting to look at the overall odds of getting picked by number of consecutive year attempted. if someone is not picked after 3 years, they've essentially had 3 tickets + 2 tickets + 1 ticket in different lotteries. Maybe the number of tickets should increase more steeply after 3 tries, or else there will be a handful of individuals with 6 tickets or more.

BRAIN TEASER for you guys ...... Does it matter if you have one or three tickets ........ NOPE .... why???

If there is only 300 slots available and 2000 people want one the odds of getting one on any selection out of the hat during the drawing is really somewhere between 1 in 3000 and 1 in 2000 (since that is your ticket odds for that selection out of the hat)..... why??

Depending on where you are in the selection process the true value of a ticket is 1 in roughly 3000 to 2000 so it really doesn't matter if you have one or three tickets. (with the first draw out of the hat each of your tickets has a 1 in 3000 chance of being drawn; with the second draw out of the hat your chance is 1 in 2999 for each ticket. (unless of course someone with more tickets is picked); with the 400th draw out of the hat your chance is 1 in 2599 for each ticket (unless of course someone with more tickets is picked somewhere along the way).

Also keep in mind the pool goes down by one as the number of slots are filled. So you only have 300 chances of being selected with 1 in 3000 (roughly) odds.

Even if you were to give 2nd or 3rd time applicants an Exponential number of tickets it wouldn't matter because then each ticket would be worth even less (1 in 100,000 for instance) so who cares if you have 100 tickets? ....... your odds of having your ticket pulled are actually worse if this was done since you would have 100 tickets with 1:100,000 odds.

So what is the underlying thought? 10% of those who apply get in and it really doesn't matter if you are a FIRST or THIRD or TENTH time applicant becuase of the ticket odds ....... the only TRUE way to increase your odds is to increase the number runners selected NOT the number of tickets in the pool ...... sorry second and third time applicants but your odds are only 0.033% better than the first time applicant on any of the 300 pulls out of the hat ......

Maybe a split lottery would be more fair. For instance if you give 50% of the available slots to the people with the most tickets (in this case third timers), 25% to the second most (in this case second timers) and the remaining 25% to the remaining applicants. Inevitably you would get drawn. If for some reason there wasn't enough applicants for one of the individual pools you could let those tickets slide down to the next pool. This would give those who have been trying the longest the best odds. The way it is set up now you could apply for 20 years and never see the start line with you own bib.

Actually, right now there are just shy of 3000 tickets. If you have three tickets, your chances of getting picked on the first draw are 1 in 999. Thereafter it gets more complicated. Each time there is a draw, an average of 1 1/2 tickets are removed from the hat. by my calculation. This is because, for instance, if a person with three tickets is picked, all three of those tickets are essentially removed from the hat. Using a spreadsheet, I calculated my cumulative odds as 0.31 (nearly one in three) assuming 60 people get special consideration leaving only 340 names to be drawn. That's a lot better than 11% for a first time applicant Credit to my brother Roy for figuring this all out.. The website cited above is not quite accurate for WS. The formula doesn't account for extra removed tickets - on that webpage, a person could have multiple winning tickets.

I think your odds of being drawn are 3 in 3000 on the first draw if you have three tickets since there is one pot of 3000 and you have 3 tickets

Right, 1 in 1000.

Hey Greg, will you guys be verifying "ticket count", as well? I hate to "narc", but I see some names of competitors from 2011 that are in the "triple ticket" count. Thanks! :)

Historically, the board seems to select international runners who are not selected in the lottery to help the global distribution of participants. For example, if there are no runners from Asia selected in the lottery, the board may decide to select a few runners from Asian countries to help the distribution. This is all based on past behavior and is not meant to predict the process going forward. If you think the board may follow past behavior, you could look at the global distribution of runners and roughly predict the odds.

All will be taken care of after the lottery closes.

If you are picked for the 2012 race you would go back to one ticket if you apply for 2013. Good luck in the lottery!

You will just have to finish in the top 10!

Have look at the stats page:
Ticket Count

The ticket count doesn't give any information on previous completion of WS100, but I thought this was an interesting question. I took a random sample of 51 names from the applicants list (when it contained 1223 names) by picking every sixth name from the first column. Of these 11 have finished WS at least once, i.e. 21.6%. So less than a quarter of the applicants have a WS finish.

Thanks Ian

Another stats page item, the Pie Chart for geographic distribution by state may not be correct, as I do not see Nevada on there, and there are Nevada runners.

You have to change the region. Click the "West" region and you will see the states in the West region. There are 21 applicants from Nevada.

To follow up on the question Doone asked — of the eleven previous WS finishers in my little sample, 7 had one prior finish, one had two finishes, one had three finishes, one had 7 finishes and one had 8 finishes. So roughly 14% have one finish, and 8% have more than one finish.

What else are you gonna do when you are injurred for years on end in this sport. Speaking of, I am finally turning the corner and have been getting in some runs. I am staying up in Elk Grove lately so let's get out there and run the flat lands. I'm free any day.

Yep, all will be cleaned after the close.

More than last year, thanks to the end of the two time loser rule. Not sure what the final count will be until we do the numbers. I'll try to post the number once the lottery closes and I have a littel time to figure it out.

A reasonable guess is that there will be about 340 places available for lottery applicants after the elite runners, raffle winners, select foreign entrants and other special categories are accommodated. At the time of writing this, there are 1398 applicants for WS 2012. Adding in the extra tickets for two and three-time applicants makes a total of 1957 tickets so far. If there are, say, 2500 applicants by the time the window closes, and the same ratio of tickets to applicants holds (1.4 : 1.0), then there would be 3500 tickets in the drawing. This would give a first-time applicant such as yourself around a one in 10 or one in 11 chance of getting selected. Lots of assumptions here, but I expect this is ballpark correct. If my estimate of total applicants if too high, then the odds would be a little lower.

They will be pulled prior to the drawing.