In the USCF rules, there are examples of players getting portions of different prizes for which they are eligible, e.g. 1/4 each of 4 different prizes for which they tied.
In the National Open this year, I got 1/4 of the U2400 prize, but I was told that I was not entitled to (up to) 3/4 of the place prize for which I tied.
A) Is this really correct?
B) If so, how is this logical or fair?
To answer A), please know what you’re talking about.
The prize distribution was correct. There were 14 players tied for 9th-17th, top U2400, and top U2300, four of whom were U2400 and one U2300. In such a case, you perform two calculations: a) add up all the prizes and divide them equally, and b) add up the class prizes only and divide them among the eligible class players. The class players get the larger of the two, which in this case was a four-way split of the U2400 prize. (And the U2300 took his class prize clear.) They don’t get that and a slice of the place prizes as well.
BTW, this really belongs under “Tournaments.”
It’s becoming obvious to me that this is the way things have always been done, but there are no examples in the rules to directly support this.
More importantly, I’m no closer to having the answer to B), and no, I don’t need the math explained again.
The text and examples in the prize distribution section seem to be chosen with two simple and logical ideas in mind:
It is still not clear to me why these principles should be violated.
It’s rule as explained above by Mr. Hillery is given under 32B3 on pages 181-182 of the 5th edition rulebook. The examples don’t give the exact case that you are dealing with (as they are examples and hence not exhaustive), but example 2 on page 183 demonstrates the basic principle.
Could you explain why you think they were violated? They don’t appear to be violated to me. Possibly you are ignoring the “up to” and are assuming that each player gets “one prize”. (That would explain your use of 3/4 in the original post.) Of course, that can’t happen for everyone when there are more tied players than there are prizes for which they are eligible. Could you give the distribution that you think would be correct?
There is also a third principle,
3) Those tied players pooled together shall have all their prizes pooled and distributed equally.
To apply the third principle slavishly (as is being done here) is to misunderstand its intention, which is that no one gets more than one prize. The reason to pool a class prize is that otherwise the player isn’t entitled to anything else. He must give up some of the class prize to get some of the place prize because he already has a whole prize and can’t get more.
In my case, the class prize is 1/4 prize, and a full share of the place prizes would 7/13 of a prize. As these add up to less than one, I should get both.
As you have pointed out, the examples in the rulebook are not exhaustive, so some extrapolation is necessary. I believe that my extrapolation is the only correct one, but in any case, if it is anyone’s belief that my interpretation is not supported by the rules, it should be.
It seems clear from 32B3 that your interpretation is not correct. Though the examples are not exhaustive, the rule is specific so there is no need to extrapolate. In addition, take a look at example 3 on page 193:
To be distributed:
2nd U2000 $100
2nd U1800 $100
Players and scores:
1675 4.5
1920 4.5
1700 4.5
1845 4.5
The rulebook gives $50 for each - pooling both prizes and dividing equally among the 4 players.
Under your interpretation, it should be:
1675: $75 (1/2 of U1800 and 1/4 of U2000, which adds up to less than 1)
1920: $25 (1/4 of U2000)
1700: $75
1845: $25
So, your interpretation is not correct and if you think the rule should be changed, then you should contact the rules committee about that.
32B3 cited above explicitly states:
If winners of different prizes tie with each other, all the cash prizes involved shall be summed and divided equally among the tied winners unless any of the winners would receive more money by winning or dividing only a particular prize for which the others in the tie are ineligible. No more than one cash prize shall go into the pool for each winner.
The prize split done meets that rule exactly. If you want a different split then you need to petition to have the rule changed (which I think is what your are aiming for). If the rule is changed then only future awarding of prize funds would be affected, not any from prior to the rule being changed.
As far as the fairness of the rule goes, think of it as having three groups.
The U2300 group had the option of getting just the U2300 prize or of pooling that prize in with others, joining that pool, and getting an even share of the pool. The U2300 group limited itself to just the U2300 prize.
The open group only had the option of pooling the prizes and sharing evenly in that pool.
The U2400 group had the option of getting just a share of the pooled U2400 prizes and splitting that pool with just the U2400 players, or pooling the U2400 prizes with the open group and joining that open group as even participants of the pool. Since joining an U2400-only pool garnered less money per U2400 player than joining the open pool, the open pool was joined.
As far as the 1/4 + 7/13 argument goes, you can look at the situation where the four players use 1/4 of their prize-earning potential to share the U2400 prize. Then each of them have 3/4 of their potential remaining and use that to join the pool of open players. Now there are 9 open players that have their full potential and 4 U2400 players that have 3/4 of their potential (9 + 4 * 3/4 = 12 players) to share the 7 open prizes. So each of those 4 U2400 players gets 3/4 * 1/12 = 1/16 of the open pool of 7 prizes while the other 9 players each get 1/12 of the pool of 7 prizes. Otherwise you have players wondering why they get less than an equal share of the prize fund they are pooled into.
The “fractional prize” method Mr. Schuyler seems to be looking for has been used in a few areas (I believe it was popular in the Northwest at one time), but it has never been widespread and it is definitely a rules variation. The NO prize division was clearly in accordance with the rules, and since those rules were widely available in advance, he is not entitled to complain about it now.
If, on the other hand, he wants to argue that the other system is “more just,” that’s another matter. I disagree – the class player is already getting more than other players with the same score, and I see no justice in letting them benefit further at the expense of equally deserving players – but this is a subjective question on which reasonable people may disagree. But I think it’s fair to ask of anyone making this argument whether he would take the same position if he were on the other side of the ratings cut.
If they did, wouldn’t they argue that the U2400 prize is among 5 players (including the U2300 player) and not just divisible by 4?
Yes. In the case described, there were 14 players tied, four of whom were U2400 and one U2300. Under Mr. Schuyler’s theory, the U2300 would have received his class prize clear, plus 1/5 of the U2400 (if you’re U2300, by definition you’re U2400), plus 1/14 of 9th-17th. The four U2400s would each have received 1/5 of the U2400 prized, plus 1/14 of the place prizes, and the remaining nine would have received 1/14 of the place prizes each.
An equivalent (and more useful) way of stating Mr. Schuyler’s argument is that class and place prizes should be calculated and awarded separately. This is an interesting theory, but the USCF rule is that place and class prizes should be combined and divided equally unless the class player would get more by taking his class prize only.
I was active in the Northwest from 1974 to 1997, and I don’t recall that approach being used at all, let alone being “popular”.
Erik Pearson, a player and occasional organizer from that area, claimed that it was, and tried to persuade us to use it in Massachusetts. It was tried a few times between 1974 and 1979, but proved very unpopular and was soon dropped.
His theory appears more complicated than that and would be horrible to calculate without a computer programmed to do so). He presumably wouldn’t add anything to the U2300 winner from the U2400 and Place prizes because the U2300 winner already had a full prize (i.e., 1 + 1/5 + 1/14 > 1). However if there had been two U2300s, they would each get 1/2 of the U2300 + 1/6 of the U2300 + 1/15 of the Place prizes.
Thank you Martinak and jwiewel. It’s nice to finally get answers that indicate an understanding of the question. Martinak, you are exactly correct about the example given–it explicitly contradicts my interpretation.
You probably won’t be surprised to hear that I still don’t see the logic. The second prize division seems more equitable to me. Perhaps I will take it up with the rules committee. The current implementation is not only illogical, but it reduces the relevancy of class prizes and heavily punishes ties. Note that non-class players rarely have this problem as they can usually share the next prize on the list.
As far as rfeditor goes, I find it interesting to learn that the “fractional prize” system has been in use.
On the next note, I am perfectly entitled to complain–I’m just not entitled to receive redress.
I don’t appreciate the implication that I’m thinking with my wallet. It’s both insulting and irrelevant to the discussion. I will refrain from making the analogous ad hominem attack.
As far as equitable goes, the distribution system I agree with is structured so that it is class players that usually benefit from class prizes. Surely this is the whole point of class prizes. If you don’t like them, don’t have them at all. Under the current system, it is easy to construct realistic examples where class players benefit by 10% or less of the value of the class prize.
Before you ask, here’s one:
1st $200
2nd $100
3d $50
1st U2000 $60
player 1, 2100, 5 pts
player 2, 2100, 5 pts
player 3, 1900, 5 pts
player 4, 2100, 4.5 pts
player 5, 1900, 4 pts
“Normal” way: p’s 1, 2, 3, (200+100+60)/3=$120, p4 $50
p5 gets nothing.
“Fractional” way: p’s 1, 2, $116.66, p3 $120, p4 $16.66, p5 $40
Of course, the idea is not to compare one to the other, but to compare each to what would have happened had there been no U2000 prize. In both systems p’s 1, 2, + 3 get $166.66.
Thus in the “Normal” way, the A-player (3) gets $3.33 extra and the rest of the $60 goes to improving the open prizes, most of it to p4.
In the “Fractional” way, $43.33 of the $60 goes towards A-players’ prizes.
My point is not that class players are fundamentally more deserving than others, it is that if there are class prizes, they should benefit class players (within reasonable limits) not go into the general pool at every plausible excuse.
So 1 and 2 each get 1/3 of first + second + third, 3 gets 1/3 of first + second + U2000, 4 gets the remaining 1/3 of third and 5 gets the remaining 2/3 of U2000. 1-3 split three prizes by splitting all of first and second, 2/3 of third and 1/3 of U2000.
If you have 1st = 500, 2nd = 400, 3rd = 300, 4th = 200, 5th = 100, U2400 = 75 and U2300 = 75 then a 14-way tie between 10 > 2400, 4 23xx and 1 < 2300 would have the 2300 getting 1/15 of 1st-5th (5/14 prize), 1/5 of U2400 and 7/15 of U2300 (total of 150 adn 15/15 prizes), leaving 8/15 (40) of U2300 for others. The U2400s would get 1/15 of 1st-5th (5/15 prizes) and 1/5 of U2400 (total of 115 and 8/15 prizes), using up all of the U2400 prize. The other 10 would each get 1/15 of 1st-5th (100 and 5/15 prizes).
Add 6th=70 and you have U2300 still getting 150 (1/15 of U2300 is worth more than 1/15 of 6th), the U2400s getting 120 (adding 1/14 of 6th to get to 127/210 prizes) and the other 10 getting 105 (adding 1/14 of 6th to get to 17/42 prizes).
Even though it looks overly complicated to me, it is still an interesting concept.
The problem with this proposal is that the distribution of the prizes within a score group isn’t necessarily well-defined (which $100 prize a particular player claims now matters within the score group, not just for lower score groups), and will be a discontinuous function of the prize amounts.
From Chess Life, August 2008, p.6
(I retyped this by hand, my apologies if I have introduced any typos.)
Counterplay
Class prizes
Of course class players benefit from class prizes some of the time, but some of the time they do not, and it is the latter case that I wish to address. Take this hypothetical tournament prize structure and result:
1st $200
2nd $100
3rd $ 50
1st U2000 $ 60
p1 2200 5 pts
p2 2200 5 pts
p3 1900 5 pts
p4 2200 4 1/2 pts
p5 1900 4 pts
The standard prize division is as follows:
p1, p2, p3 (200+ 100+ 60)/ 3= $120
p4 $50 (3rd place)
p5 $0
What’s going on here? Where did the money go? Had there been no u2000 prize, p1, p2, and p3 would have gotten $116.66 and p4 would have gotten nothing. Of the extra $60 in the prize pool, $3.33 has benefited an u2000 player (p3), and $56.57 has benefited masters (mostly p4).
Furthermore, p5 (1900) is left in the absurd and illogical situation of wishing the U2000 prize were smaller. If it were $1 less than the 3rd place prize, he would have received it.
These situations are more common than many people realize, and they are mostly unnecessary. The solution is simply to award the appropriate fraction of each prize to each player (with the stipulation that the fractions not add up to more than one).
Under this system, the prizes would be distributed as follows:
p1, 2 1/3(200)+ 1/3(100)+ 1/3(50)= $116.66
p3 1/3(200)+ 1/3(100)+ 1/3(60)= $120.00
p4 1/3(50) = $16.66
p5 2/3(60) = $40.00
In my discussions with players and directors, I have encountered three objections to this proposal.
I don’t think that is the case, but even if it were, it’s hard to believe that that would be a serious obstacle in this day and age.
This objection makes no sense to me since it approximately describes the fundamental idea of class prizes. One might as well object to having class prizes in the first place. It is worth pointing out that a place prize winner will always receive the same or more money under this proposed “fractional” distribution system than they would if the class prizes were not in the prize pool.
I cannot vouch for either one of these statements, but I can wonder whether anyone has legitimate cause for complaint. Do top players really think they should reap the rewards of class prizes? Perhaps the current system would be unpopular if it were fully understood by all players.
–James Schuyler
Las Vegas, Nevada
[Chess Life Editor: ] Chess Life asked 2008 USCF TD of the Year Frank K. Berry to respond:
I think a discussion on theoretical innovative prize fund distributions should be done on the USCF forums as uschess.org. And I see where Mr. Schuyler has instigated on June 11 a discussion on this question in the USCF forums → Chess Tournaments → subject: “Splitting Money Prizes.”
I think the USCF rulebook holds up well in this area. Of course exceptions can be devised to show inequities. I think most class players who do score well win a prize. Especially if the class prize is designed to be split. In Oklahoma we will split a $300 class prize $150-100-50. An exception was the recent qualifier to the 2008 FKB Championships.
My problem with Mr. Schuyler’s new plan is that–yes–it is quite complicated. Towards the end of an event most players are in a hurry to have the class prize split computed, paid out and then take off for the drive home. We figure the final class prize breakdowns (using USCF guidelines) on the pre-last-round wall chart print out by constantly updating it from the results being posted on the last round pairings sheet. We can have the split on the final class prizes computed–by hand–just as the last game of each section is finishing if not before.
The USCF has larger problems than a possible new format for prize fund distribution. Mr. Schyler’s considerable talents for innovation theory would better serve the USCF on one of their committees. I hereby nominate him for chairman of the Finance committee.
–Frank K. Berry, ANDT, IA, IO
Chess Life welcomes letters from its readers. Letters are subject to editing for content and length. Send your letters to letters@uschess.org, and include your full name and a telephone number. Senders should not expect a personal response.
[advert for 2008 SPICE Cup Festival, SPICE.ttu.edu]
HUH? p1, p2 & p3 share the top three spots. And should be getting (200+100+50)/3 = $116.66
p4 is in 3rd place? Looks like he’s in FOURTH place to me. 4th place = $0
p5 is in limbo here simply because you haven’t given the parameters for the tournament
p5 may or may not get the $60 for U/2000
& p4 might be in 3rd place and get $116.66 & p3 would get the U/2000 prize, depending upon the parameters for the tournament
And either p3 should get the U/2000 $60 or p5 should. This depends upon how the tourney was set up originally. If the tournament was set up that the U/2000 players only played against other U/2000 while everyone else either played in the “Open” section (mandatory for 2000 and above players), then it would depend upon which section p3 played. OR if it was a totally open tourney where a 1600 player could end up against a 2200 player, then this would have to be spelled out. p3 could possibly be able to get $116.66 for finishing in top place plus the $60 for best U/2000. And whether or not players are eligible for more than one prize would have to be spelled out.
Simply put. Give the parameters for the tournament. Is this tournament completely open? Or are there class divisions that only those that have ratings allow them to play in? I once played in a tournament where I played only those in my class/division. I had a 5.0 score. I ONLY received the class prize and nothing of the top prize. (I think the top player in the tournament had a 4.5 score. His rating was around 2200 I think.)
Depending on the circumstances of the tournament. Your numbers could change radically. To wit:
p1 = $116.66
p2 = $116.66
p3 = $116.66
p4 = $0
p5 = $60
or
p1 = (200+100)/2 = $150
p2 = (200+100)/2 = $150
p3 = $60
p4 = $50
p5 = $0
p1 = $
However, rule 32B1 states:
One cash prize per player. No winner shall receive more than one cash award. etc.
Therefore, either p3 receives 1/3 of the top cash award ($116.66) or the U/2000 ($60). Not both.
And one last thing, how did, in your example, the $60 U/2000 award get mixed into the top three player award money rather than the $50 for third place? You are awarding $20 of the U/2000 money to two players over 2000?
Rule 32B1 states, in part, “A clear winner of more than one cash prize must be awarded the most valuable prize.”
Rule 32B3 says “If winners of different prizes tie with each other, all the cash prizes involved shall be summed and divided equally among the tied winners unless any of the winners would receive more money by winning or dividing only a particular prize for which the others are ineligible. No more than one cash prize shall go into the pool for each winner.”
Thus, per 32B1, the three prizes summed up for the top group include the more valuable $60 U/2000 instead of the $50 3rd place.
Per 32B3, the $200, $100 and $60 prizes are split among the top three, leaving the $50 3rd place prize for the next highest scorer after those three.
I think James Schuyler’s proposed distribution has some merit, but … Once class prizes are recognized as weird and illogical themselves, it is hard to get worked up about whether they are divided logically.
As a class player coming up I won plenty of class prizes, and although I understood they were not “real” prizes I was always very pleased to get them. As an organizer I had to divide up scant money, and at first I agonized over what the place-class split should be, but I learned it doesn’t really matter that much.
About the only class prize I care about nowadays is the U2400 prize, but not the way you might think. (1) Given all the stronger-than-me players around, in a big multi-section tournament where I have to enter the top section, the U2400 prize is the only one I have a decent shot at. But you might be surprised to learn that I don’t look at the prizes offered before entering a tournament. Even big money tournaments give a negative $ return over the long haul, so I basically consider all the costs and plan on $0 as my prize. Other than that I focus strictly on the chess. (2) If at the end I am in contention for a prize then I try for it – I’m not crazy! But the money, if I succeed, is really a bonus for a good tournament. (3) There shouldn’t be any U2400 prizes! U2400 is not a class – there are IMs in that group! All that money should be in the place prizes.
About the best argument for class prizes I ever heard was from a gentleman named Tom Kane, approximately 1400 rating. According to Tom, he likes class prizes because that means that in the last round his opponents are really trying. Well put, and works for me too (see point 2 above).
Now to the specific points put by James Schuyler in Chess Life:
What’s going on is that there is no such thing as “extra” money when prizes are funded by entry fees. The alternative to $200-100-50 U2000 $60 is not $200-100-50 but instead $230-120-60. The so-called “extra” $60 class prize is formed by taking away $60 from the place prizes (high scores) to reward the class player (high score with arbitrary low rating). So when constructing your alternative prize scenarios, please try to limit yourself to choices that an organizer would reasonably make.
One way to look at it is that p3 is now a place player, so p5 becomes the top U2000 and gets that prize. Only if p3 could do better by taking the class prize she should do that, and the remaining prizes divide up without p3 and her class prize. Or if p3 could do better by including the class prize in the division then she should do that, and again the remaining prizes divide up without p3 and her class prize.
That is pretty much what the current rules do, and I don’t have much objection to that.
Absurd (and amusing), but not illogical. Organizers just need to be aware that having class prizes slightly larger than place prizes will sometimes lead to the class prize apparently going to a non-class player. In fact I thought it was common knowledge that in order to minimize arguments, the class prizes should start at slightly less than the last place prize (in the same section), and ALL the prizes should be different dollar amounts. In the example given, because the $50 and $60 prizes are similar in size, it made only a small difference to p3 which one he got, but it made a big difference to p5. So psychologically we tend to think that p4 got the class prize, but that didn’t really happen. If the prize fund had been $175-75-25 U2000 $135 then p5 would also receive nothing under current rules, but for some reason that seems fairer. Probably the least confusing way to construct class prizes would be like this: U2000 $200 1st-3rd $100-60-50. Then there would never be any arguments about how the money was split. (Just kidding on this last point, obviously.)
Not necessarily better, just different. And to show how easy it is to label something as “illogical” – under the proposed system, p5 gets a larger share of U2000 money than p3, even though scoring fewer points! Illogical! (Just kidding again.)
The rulebook used to be ambiguous on which prizes to put in which pool, so it was then legal to distribute prizes ($200-100-55 U2000 $55) in this way: p1,2,3 $118.66 p4 $0 p5 $55. Organizers would sometimes do that to preserve the intent of the class prizes. Or sometimes they would give p4 the $55 to preserve the intent of the place prizes. Either way they would get an argument. Now there is 32B4 which means that p5 gets the prize, so obviously the rules committee agrees with you on principle that it is meant as a class prize.
An unrelated rant: Whenever the rulebook can’t cover all possible cases, it should be silent. Rules shouldn’t be written just to prevent future arguments, simply because it can never work. If someone wants to argue, then more rules just means more things to argue about. Related to this is that Chapter 2 is titled “USCF Tournament Regulations and Guidelines”. Are they regulations or are they guidelines? Come on folks! If they are not binding then they belong in a different place, otherwise you are just inviting arguments.
The reason this is a serious objection is that players will not know what score they need when they sit down to play the last round. This is a well-known problem – the same thing is true of some tiebreak systems which are used for indivisible prizes, like scholarships, qualification spots, etc.
I witnessed this for myself at Foxwoods 2002 when a player finished his last round game, inspected the wallcharts, and said to his friend: “I’m in.” [Meaning the 2003 U.S. Championships.] I thought that was an amazing display of calculating ability [Sonneborn-Berger tiebreaks if I recall correctly] because I had already decided it was too complicated for me to figure out. Not so amazing though because he was simply wrong, it still depended on some unfinished games, and in fact he was out. At the point his game finished even a computer could only have said “it depends”. I don’t recall if he drew his last-round game, although it would make my story better.
If players are not playing for money then pretty much who cares how the prizes are split.
But when it gets to the last round, most in contention are in fact playing for money, me included. If playing for money, they have a right to know how much money they will get for a win, loss or draw. True, under the current system it also depends on the results of other games, but the decision tree is fairly small and quickly narrows as other results become known. Under the proposed system it is always murky with more than two games unfinished.
This means that the proposed system debatably improves fairness at a certain cost of clarity for all players.
Um, here we go again, considering the class prizes as “extra” prizes.
This is a weak counter argument. To heck with the players! (Kidding, of course.)
I can’t speak for what all top players think, but here is what I think: Class prizes are stupid, and I am against them even though more entries means more place money, which is what I am shooting for. I think class prizes weaken chess and the USCF in the long run, and organizers would do better to run team events, amateur events, round robins, scholastics, and place-only one-section swisses. But I enter the tournaments that are offered.
I think most players understand the current system well enough, and in their gut they agree with me that class prizes are illogical and not worth arguing over. Except when they didn’t win one they expected to.
–
2008.08.03 20:39 - edited post to remove “you or your” in 4 places