Normal prize distribution rules apply. When the dollar amount is the same then the higher status prize is the one that is awarded. In your example if a master, expert, A player, B player and C player tie for third and best U1800 then it is a five-way split of $500. If 3rd is worth more then it is still a five-way split of the two prizes. If 3rd is worth less then it is a two-way split of $200 and a 3-way split of $<300.
If 3rd is $300, best-U1800 is $300 and a 1583 is in sole 3rd place then 3rd is awarded.
PS the prize discussion is starting to be a tangent that may deserve being split off from the main thread.
What if there are only four players in the tie? Which $100 prize don’t you award to any of the four tied players? To put it another way, which player(s) in the lower score group(s) get shafted out of the fifth $100?
All four players get $125 and the lower score groups are shut out of those prizes rather than shafting the four tied players out of the $100 that should also be split between them. Whether it is a tie with two players, three players, four players or five players, the two prizes are split with no need for any to be reserved for yet another player. Though if none of the tied players was eligible for the U1800 prize then that one would be (entirely) unawarded to the tied players and available for others.
This discussion is evolving how I thought it might.
There are several deficiencies in the form of things where reasonable people might interpret the language differently.
Most of those deficiencies ARE handled by the rules.
Understanding of how those deficiencies are addressed by the rules isn’t uniform, even among the various rule moguls and rule mogul wannabees on these illustrious forums.
If someone just told the organizers for the Washington Chess Federation that this flyer indicated poor practices, they wouldn’t have a clue what the fuss was about.
The TPR rather than score is just weird. Yes, it’s possible that someone who is 3-3 to have a “better” tournament than someone who is 3.5-2.5, but then it’s possible for someone to win bigger place money by executing the perfect Swiss gambit over someone who starts with a string of wins and loses games late. We live with it.
“Best Female Player: $100, $50, 1-yr NWC
subscription extension. Awarded to the best scoring
(by TPR) female players who do not win another cash
prize. Requires min 3 qualifying players per section”
“do not win ANOTHER cash prize?” Shouldn’t that be HIGHER cash prize? And (as described) it should be handled by being distributed with the other section money prizes; it’s not an additional prize pool like the best game.
How is this “clear”? The only section that even mentions Unrated is the bottom one. How can you argue that on the one hand, unrateds aren’t eligible for U1500 money because it doesn’t say U1500/unrated, but that they ARE eligible for U1800 section money even though it doesn’t say U1800/unrated? (BTW, that’s rhetorical). The language (or lack thereof) regarding unrateds in this is just dreadful. It looks like the intent was to limit unrateds to just the unrated money in the U1400 section (other than place money in the open, for which they are always eligible), but they misfired on that, badly. Similarly the language regarding the female prizes is flatly wrong (and that’s ignoring the odd use of TPR rather than score).
I disagree with Allen about class vs under prizes, though the point that if you are going to use under prizes, it’s a good idea to have them slightly higher for the higher ratings—it does make it easier to explain a high-scorer U1900 taking the U2100 money.
Using TPR to award prizes is seldom a good idea, because TPR depends on what ratings you use: published vs latest (eg, from MSA), the former possibly being non-representational of the player’s current strength and the latter being subject to temporal changes. Also, there can be more than one way to compute TPR, especially when they play unrated players.
I’m sure not staying the age requirement to be eligible for this was just an oversight on the flyer. I don’t think the flyer indicating that you can play in the main event and side events was referring to the adult-only swiss.
But this is my point. This is a poorly designed flyer and, by extension, tournament. Surely you don’t think this flyer is exempt from being considered part of “pre-event publicity.”
But this is impossible. If there is a $200 prize for U2000 and a $200 prize for U1800 and there is an A player with 4.5 and B players with 5 and 4.5 and all other players winning place prizes or having fewer points then the B player with 4.5 gets more money if the B player with 5 “chooses” the U2000 prize, but the A player gets none at all. I see that this doesn’t answer your question exactly, but perhaps it leads you to where Mr. Priest has gone.
Here is (I hope) a better one: Suppose the following prizes are all announced:
1st overall — $500
2nd overall – $400
3rd overall – $300
4th overall – $200
5th overall – $100
1st U1800 — $250
2nd U1800 – $100
– and there are 8 players with a score of 3.5 or better:
2000 - 5.0
1900 - 4.5
1740 - 4.0
1730 - 4.0
1720 - 4.0
1710 - 4.0
1850 - 3.5
1750 - 3.5
1st and 2nd are easy.
Following are 4 players, tied, all under 1800. So the 4 highest remaining prizes – 3rd overall, 4th overall, 5th overall, and 1st U1800 – are combined and split equally among these 4 players.
Finally, there are 2 prizes remaining, each $100, to be awarded (in some fashion) to the 2 players with 3.5. Which of the two $100 prizes do we pay out first? Why does it make any difference? They’re both $100.
Well, if we first pay the 5th overall, then the 2nd U1800 goes exclusively to the player with 3.5 who is under 1800. But if we first pay the 2nd under 1800, then the 5th overall gets split evenly between the two players at 3.5. So it makes a big difference for these two players.
Second question: If we change the prize fund so that 5th overall is $101 and 2nd U1600 is $99, how do the two prizes get distributed? Or, if we do it the other way around and change 5th overall to $99 and 2nd U1600 to $101, what happens now?
Tournament Performance Rating. The simplest form of this is the old US Chess “provisional” formula of averaging
opponent rating+400 on wins
opponent rating-400 on losses
opponent rating on draws
I assume that’s what they intend to use. Mathematically it reduces to C.A. + 800*(score/N-.5), where C.A. is the average rating of the competition (“competition average”) and N is the number of games. (There are a variety of more complicated alternatives. FIDE, for instance, uses a non-linear function of the percentage score: score/N). It has a number of problems:
a. What do you do when the opponent is unrated?
b. Oppo+400 for a very low rated opponent is a low number which likely isn’t very indicative of the skill level. In (e.g.) the U1400 section, someone who beats a 200 in round one will have a very difficult time overcoming the hit that makes to the TPR in the remaining rounds. For instance, if someone else beats a 900 in round one, you would need to score a full extra point in a six round tournament.
c. Oppo-400 for a loss to a high rated individual may be a number which is too high to be indicative of the skill level. That might be an issue in the open section (though much less likely to be a problem than b in the bottom section).
As my college statistics professor used to say, if a man has one foot on a block of dry ice and the other in a pot of boiling water, on average he’s comfortable. That’s the flaw in performance ratings.
I wouldn’t say very recently – we stopped doing that several years ago. But I’ll admit that it was a bad practice that should have been discontinued much sooner. Keep in mind, though, that the way we were doing it was the way everybody did it back in 1985 when I started playing rated chess. Adjournments were a standard part of the game back then. We just continued using that model longer than others did.
However, that is a completely different issue than “under” prizes vs. class prizes. Making a bad decision on one issue doesn’t mean that all our decisions have been bad, or that all our practices are bad. Multiple “under” prizes in the same section is a bad practice in my opinion, due to the reasons given above – although my language in that post might be a bit confusing. Several times I referred to multiple “under” sections when I really meant multiple prizes within a single section. A single player should not be eligible for more than two prizes (one class or “under” prize, plus one of the overall prizes (1st/2nd/3rd/etc. in the section)). And even then, he usually cannot win more than one prize unless he’s tied with other players, in which case the only restriction is that the number of combined prizes cannot exceed the number of players.
If you have multiple “under” prizes (in the same section), any player who is eligible for one of the lower ones is also eligible for all the ones higher than that, because he is “under” those rating limits as well as his “natural” one. And that can make determining the prize distribution much more complicated. Class prizes have the advantage (a great advantage in my opinion) that a player can be eligible for at most one of them.
When I calculate TPR for my own use, I use the above formulas in all cases except two:
(1) If I beat someone who is more than 400 points below me, I just use my own current rating (on the principle that you should never lose rating points by winning).
(2) If I lose to someone who is more than 400 points above me, I again just use my own current rating (on the principle that you should never gain rating points by losing).
In both cases, I have performed exactly as expected, so my rating should not change. These “cheats” are generally not necessary in high-level tournaments, because everyone is usually within 400 points of everyone else. In an open swiss, though, you can get some bizarre results if you don’t use them. If I lose to 5 GMs, it would be ridiculous to give me a TPR of 2100+ (not that I would ever play 5 GMs in an open swiss, but the principle remains: even one such result would skew the overall TPR when it gets averaged in with the others).
ETA: For unrated opponents, I generally use their post-tournament provisional ratings in my calculations.
When prizes are equal dollars you go with the higher ranked one (an overall prize is higher ranked than an equal dollar under prize, a U2000 is higher ranked than an equal dollar U1800).
The 1850 at 3.5 is out of luck.
Assuming you meant U1800 instead of U1600, the $101 prize is higher ranked than the $99 prize and is the one pulled into the four-way tie. So if the overall prize is larger (or equal) then that is the one used. If 2nd U1800 is more than 5th overall then 2nd U1800 is pulled into the four-way tie and the A and B players would evenly split 5th overall.