On a couple of occasions, I have been advantaged by having a class prize mixed in with the place prize money. Had the class prize player shared in only the place money he would have received more than he would have as the class prize winner. By moving up his prize, we higher rated players made more money by sharing in this prize. So did he, but not as much as we did by having his prize move up to be with our place prizes. The class players who could have split the class prize were left in the cold and were very bitter about it. I told the TD in one tournament that the fairest way was to split the place money among us and give out the class prize to the class players as it would include more players as winning a prize in the event, but he insisted on doing it the USCF/CCA way. We had a long discussion on the pros and cons of Rule 32 and on how the tournament was set up which caused the problem in prize distribution. The other high rated players enjoyed getting extra money but were surprised that they were getting a part of the U2000 prize, too.
When I started out in directing, I was told to try to include as many individuals in the prizes as possible and to see to it that the class players max out in the number and size of their prizes. Place prize winners won or shared in the place prize money. Class players who won a place prize would receive the maximum he was entitled to as long as it was above the class prize. Otherwise, he would receive the class prize. The remaining class prize money would go to the other class players creating more prize winners. Under the present system, the majority of class players are disadvantaged by the rules for prize distribution.
I fully understand the prize distribution system and the “logic” behind it, but having been advantaged by it, I feel it is generally unfair to lower rated players and has a bias that funnels too much money up to the top players. The lower rated players are the ones who fund these tournaments. My tendency is to be biased for the little guy whenever possible.
Please clarify. I think I followed what you said but I want to be certain.
Assume the prize fund is $1000, $700, $400, U2000 $500
I think you said that an A player tying for second would get $550 (1/2 each of 2nd and 3rd) instead of $600 (1/2 each of 2nd and U2000), while an A-player taking clear third would get $500 (U2000) instead of $400 (third).
The $550 tie for second would mean that the top U2000 would get $50 less (with a 2nd/3rd split as opposed to a 2nd/U2000 split) but would leave the full $500 U2000 prize available for the second U2000 player. That lets U2000 players as a whole take $550 more with your split even though the top player takes a $50 hit.
Next let’s look at a three-way tie for second (at 4.5-1.5) with one U2000 involved. A three-way split of 2nd/3rd/U2000 would give the players $533 each (the U2000 group as a whole gets $533). A three-way split of just 2nd and 3rd would give the players $367 each and leave the $500 U2000 for the second U2000 player (at 3.5-2.5), giving $867 in total to U2000 at the cost of the top U2000 getting $133 less than if he’d scored a half-point less (4-2). A third option would be to have the top U2000 take only $500 (more than a three-way split of 2nd and 3rd) and let the other two players split the top two prizes for $550 each (the U2000 group as a whole gets $500). A fourth option would be to pull $400 from U2000 to make a fourth place prize and have the three-way split of 2nd-4th give $500 each (so the top U2000 makes at least as much for going 4.5-1.5 as for going 4-2) and leaving $100 for the second U2000.
Boyd, you really need to un-dig your heels on this. 32B3 describes the procedure to follow in case of a tie.
There are no qualifiers on “all”. Not only is there not an implicit “only the prizes for which all players in the tie are eligible”, but in fact the “unless…” clause rather explicitly allows for the fact that the players may not be eligible for the same prizes.
If you ignored the black-letter description of how to handle a tie and someone who was shorted appealed, they would win.
The Boardwalk Open is probably a more complicated situation, because:
There are two principles involved:
a.) The application of Rule 32B3 (the most recent amendment not germane in this case).
b.) The “correct” algorithm for distribution of prizes involving a limited prize
The amount of the unrated prize limit was coincidentally the same as the amount of the second-largest prize (in this case, the 1st Under 1400 prize was the second-largest prize).
I was concerned about justifying the prize distribution amounts, and this was the explanation Bill Goichberg gave me:
The rated player brought first into the tie and the unrated brought 2nd. The unrated could not bring more than $600 into the tie, and is not eligible for U1400, so the only prize he can bring into the tie is 2nd. The rated player is the only one who can bring $1000 into the tie, and he can not also bring in U1400.
He did recommend, because the cutoff issue presents a problem, it might be a good idea to stipulate that unrateds are not pooled with rateds if:
1)The unrateds are not eligible for prizes in that class and
2) Without pooling, the unrateds in the tie will not win more than the rateds.
I also posted the above as an edit to a previous post of mine as well.
I’ve quoted Rule 32B3 in this thread. I’m also fairly sure Rule 32B1 precedes it.
The black-letter rules, as several people have already said, are a mess. So, I use the principles I’ve outlined in this thread when I must distribute limited prizes. I acknowledge my application isn’t perfectly consistent across all situations, particularly those prize funds that do not adhere to Rule 33B.
Perhaps surprisingly, I’ve done this once or twice. I’ve also defended my procedure - to players, directors and organizers.
Of course, worrying about appeals is probably one of the worst things I could do when making a decision. I make the best ruling I can, and live with that.
Again, I’ll agree to disagree. Other than that, I’ll rely on my previous posts, and likely end my participation in this thread.
Assuming, strictlyarguendo, that your above condition is in place, I think the simplest approach would be to make Rule 33B a requirement for prize fund construction, instead of a recommendation. These problems only really occur when you have a class prize that is larger than most/some of the place prizes.
(One should not infer from the above that I favor making Rule 33B a requirement, although it is one of my guiding principles when I design prize funds. I will note, though, that the previously cited Rule 32B5, Example 3, does conform to Rule 33B.)
Nor do probably 80% of all Grand Prix tournaments including almost all GP tournaments (CCA or no) that offer 30 or more GP points. You can’t follow 33B unless you either don’t go very deep on overall prizes or your under prizes are rather thin.
Besides even a 33B compliant prize fund doesn’t avoid “problems”. Suppose you have overall prizes of 500, 250, 150, 100 and U2200 prizes of 75 and 50. Assume A, B and C tie at 4.5 and D (over 2200) and E, F (both under) tie at 4.0. I defy you to come up with any reasonable way to handle the 4.0’s other than giving each of the three players ($100+$75+$50)/3. And if you decide that $75 each is correct, I defy you to come up with an explanation which involves D getting none of the under money and E and F getting no more than a total weight of 1.0 on the three prizes.
That’s an interesting principle—unfortunately
It isn’t part of the rules and in fact is explicitly rejected in 32B3.
It can’t be used as the basis for a coherent strategy for distributing money (as in the example here, the example in the rulebook, etc.).
I think we can all agree that the rulebook language regarding the handling of players with prize limits is unsatisfactory—as Bob mentions above, it could be used to justify a shockingly wide range of “solutions.” But the prize distribution where no one has an external limit should be 100% reproducible (and Bob’s laudable goal is to extend that to situations with an external limit). If there’s a problem with the existing 32B1 and 32B3, then we should probably endeavor to fix them, since the limit calculations are based upon those.
Boyd had earlier said that he would bring in the prizes in sequence starting with the place prizes. Since first through third are already allocated he can stay consistent and still bring in fourth and both U2200 prizes for the three-way split.
That consistency would mean that a $175 top U2200, $125 fourth and $25 second U2200 would still have the same three-way split for first even if a U2200 was in the tie and the prizes would be brought in for the fourth-sixth tie in the sequence $125, $175, $25.
I think I am understanding him correctly, but he can correct me if I’m not.
This is no different, conceptually, than Example 3 of Rule 32B5, the distribution of which I’ve already said I would use (of course, I’ve also said why I would use it).
Yet, that same concept was used in my analysis of the Boardwalk Open distribution quoted in this very thread - which, IMO, is an excellent practical example of the problems with a prize fund that is not compliant with Rule 33B. I note that Mr. Goichberg’s justification for his decision, as quoted by Mr. Immitt, uses this concept as well.
I’ll pass. I’m not sufficiently motivated to flog this particular horse, especially for a situation that I have only encountered at CCA events, where the final decision/approval on limited prize distribution is never mine, anyway.
It makes sense that Mr. Immitt, Mr. Messenger and Mr. Hater (who, combined, are involved with running most CCA events onsite) want clarification on this, and I’m fine with following whatever rules change may be made.
When I took the senior TD test there were several questions on prize distribution each with four answers* giving different prize distributions. With rule book in hand, I never calculated total agreement with any of the proposed answers. Two of the answers were always clearly wrong, so I guessed from the remaining two plausible answers.
This was a Great Personal Failing. But when I calculated prize distribution at actual tournaments, no one ever complained. Of course, the largest total fund I ever worked with was about $2000.
I remember the Senior TD test as being four answer multiple choice questions. I could be wrong; it was a long, long time ago.
The “No player shall be awarded…” may not be the best way to handle that in situations where there are under prizes mixed in. Take the Boardwalk situation, but change the limit from the serendipitous $600 to $533. Under the proposed rule, Aaron would get $1000, Kyuso $533, but the balance would have to flow past the 4 and 3.5 groups since the first person who would have a different amount without the limited player would be the U1400 down at 3.0. Not the worst possible outcome, but if the limit were like $200, you would still have to run the extra money all the way down to 3.0. Alternatively, if you use the concept that coalitions should maximize their combined take, Aaron, Kyuso and Whiteside combined would bring in $1000, $600, $533 which can be distributed $1000, $533, $600. With a $200 limit, you would have Aaron $1000, Whiteside $600, then you would have the $533, $267 and $133 available to Kyuso and the two 3.5’s. Kyuso would take $200 and the others would split the remaining $733. The money spreads out among all score groups down until you hit a prize <= the limit (possibly 0 as in the OP).
WinTD does, though it doesn’t have any method of handling the limited prizes discussed here (since, as is rather clear, there is no established procedure for doing them). See
As I see it, Whiteside shouldn’t be able to win more than $533 because that’s what he would have received for clear second if Kyosu were excluded from prizes. If Aaron (limited to $1000) and Kyosu (limited to $533) pool the 1st + Top U1400 prizes, with $67 dropping down to the next player in line, you’re right that it would skip several places under the proposed rule since Whiteside would be limited to $533 and Ming would be limited to $267. I think that’s better than letting Whiteside win $600.
Under the CCA system, Aaron and Kyosu (limited to $533) would split 1st and 2nd (Aaron $1000, Kyosu $533), with no money dropping down. The remaining prizes would be awarded as they were in the actual tournament.
If Kyosu had a prize limit of $200, under the CCA system we’d first award Kyosu his prize of $200, taken out of the $267 3rd prize, with $1,000 going to Aaron. $67 would drop down from 3rd. Whiteside would get the 2nd prize of $533, and Ming would get $133 for 4th plus $67 dropped down for a total of $200, which is less than his limit of $267. Culp would get $600 for Top U1400.
That’s Bill Goichberg’s method: first deciding which prize the limited prize comes out of, in this case 3rd prize. If you start at the top it produces the same result, assuming the U1400 prize is out of the picture. Aaron and Kyosu win $1000 + $533, but Kyosu can only win $200 and Aaron can’t win more than $1000 so $333 drops down. Whiteside wins the 3rd prize of $267 plus $333 dropped down = $600, but he’s limited to $533 so $67 drops down. Ming wins $133 plus $67 dropped down = $200.
I’m trying to figure out how to program this and the process of dropping the limited player(s) down and carrying money with them seems to be more generally applicable. The (*)'s in the calculations are prizes brought down with a player who is entitled to his limit.
Prizes
Open
$1000
$533
$267
$133
Under
$600
Results
5.0 ($533 limit)
5.0 (no limit, under)
4.5 (no limit)
4.0 (no limit)
4.0 (no limit)
4.0 (no limit)
5.0 score group
$1000 (1st) + $600 (Under). Give $1000 to Aaron. Remainder is too big for Kyuso, so carry $600 down.
4.5 score group
$600(*)+$533. $533 to Kyuso(5.0), $600 to Whiteside (4.5).
4.0 score group.
Standard
Same thing with $200 limit
5.0
$1000 (1st) + $600 (Under). Give $1000 to Aaron. Remainder is too big for Kyuso, so carry $600 down.
4.5
$600(*)+$533. Give $600 to Whiteside. Remainder is too big, so carry $533 down.
4.0
$533(*)+$267+$133. Give $200 to Kyuso. Split remaining $733 among the other two.
I tried this with a couple of other rather messy distributions discussed on the forum and the idea seems to work. Again, it’s based upon the idea that coalitions of the top scorers are entitled to maximize their collective prizes which is what drives 32B3—it’s just that you have to extend that to coalitions that span score groups. (Of course, if there are no limits, if each score group maxes out from top to bottom, the coalitions of top scores also per force do that). Note that you never actually create a new prize—you just drift prizes down with players until you hit a spot where you can take out their whole share. It will give the same result as the CCA method if there are just “overall” prizes, but not (necessarily) if there are under prizes.
That’s not how I’d do it. In a CCA tournament the 5.0 score group would get $1000 (1st) + $533 (2nd), but since CCA tournaments don’t use WinTD that’s probably not a major consideration for you. It would be better to ask the chief TDs and pairings chiefs from the last few U.S. Opens and National Opens how they’d do it. I know Boyd Reed would give $1000 (1st) to Aaron and $533 (2nd) to Kyosu. Maybe you could add one or more options to let the pairings chief decide which method to use.
Before 2014 I would have given $1000 (1st) plus $600 (Under) to the 5.0 group for a total of $1600, then applied the prize limits, giving Aaron $1000 (can’t win more than clear 1st) and Kyosu $533 (his limit). Now I’m becoming sympathetic to the idea that a player with a prize limit shouldn’t be able to pull more than that limit into a prize pool, which would mean giving $1000 (1st) to Aaron and $533 (2nd) to Kyosu, since that gives them more money than if Aaron pulled in $600 (Under) and Kyosu pulled in $533 of the $1000 1st prize. But since out of the tournaments which I direct these situations mostly come up in CCA events, I think I’ll just use the CCA method from now on, even in MACA tournaments.
In the CCA method Kysosu’s $200 would come out of the $267 third prize, with Aaron winning $1000. A top-down version of the CCA method would give the 5.0 group $1000 (1st) plus $533 (2nd) for a total of $1533. Aaron would win $1000 (can’t win more than clear 1st), Kyosu would win $200 (his limit) and $333 would drop down. Culp would win the $600 Under prize.
So you see $600(*) + $533 as being two prizes, and Whiteside can only win one of them? In a top-down CCA, with $333 dropping down from the 5.0 score group and $267 for 3rd, I’d add the two amounts, giving the 4.5 group $600, but Whiteside would be limited to $533 so he’d get that and $67 would drop down.
That’s not how I’d do it. In a CCA tournament the 5.0 score group would get $1000 (1st) + $533 (2nd), but since CCA tournaments don’t use WinTD that’s probably not a major consideration for you. It would be better to ask the chief TDs and pairings chiefs from the last few U.S. Opens and National Opens how they’d do it. I know Boyd Reed would give $1000 (1st) to Aaron and $533 (2nd) to Kyosu. Maybe you could add one or more options to let the pairings chief decide which method to use.
Before 2014 I would have given $1000 (1st) plus $600 (Under) to the 5.0 group for a total of $1600, then applied the prize limits, giving Aaron $1000 (can’t win more than clear 1st) and Kyosu $533 (his limit). Now I’m becoming sympathetic to the idea that a player with a prize limit shouldn’t be able to pull more than that limit into a prize pool, which would mean giving $1000 (1st) to Aaron and $533 (2nd) to Kyosu, since that gives them more money than if Aaron pulled in $600 (Under) and Kyosu pulled in $533 of the $1000 1st prize. But since out of the tournaments which I direct these situations mostly come up in CCA events, I think I’ll just use the CCA method from now on, even in MACA tournaments.
In the CCA method Kyosu’s $200 would come out of the $267 third prize, with Aaron winning $1000. A top-down version of the CCA method would give the 5.0 group $1000 (1st) plus $533 (2nd) for a total of $1533. Aaron would win $1000 (can’t win more than clear 1st), Kyosu would win $200 (his limit) and $333 would drop down. Culp would win the $600 Under prize.
So you see $600(*) + $533 as being two prizes, and Whiteside can only win one of them? In a top-down CCA, with $333 dropping down from the 5.0 score group and $267 for 3rd, I’d add the two amounts, giving the 4.5 group $600, but Whiteside would be limited to $533 so he’d get that and $67 would drop down.