|
Post by mikegarrison on Oct 24, 2014 10:34:54 GMT -5
Besides, everyone on here uses set scores and game scores to argue who is better anyways. Why not use it? Actually, many people don't care about set scores or number of sets, just wins and losses. Pablo is based on the number of points scored. IIRC, bofa did a lot of work on including set wins but found that they didn't really add a significant increase in predictive value compared to the total points scored in the match. Perhaps he could comment.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 24, 2014 11:07:32 GMT -5
I thought this was an interesting topic. If you were to apply this to the B1G conference standings this is how it would look (assuming my quick math is correct): 1- Wisconsin (24 pts) (7/3-0 wins, 1/3-1 win)2- Purdue (20 pts)3- Illinois (18 pts) 3- Nebraska (18 pts) 5- Penn State (17 pts) 5- Michigan (17 pts) 7- Michigan State (15 pts) 7- Ohio State (15 pts) (2/3-0 wins, 1/3-1 win, 2/3-2 wins, 2/2-3 loss)9- Northwestern (9 pts) 9- Minnesota (9 pts) 11- Indiana (8 pts) 12- Iowa (3 pts) 12- Maryland (3 pts) 14- Rutgers (1 pt) I think the hardest part of accepting this formula is it's non-traditional. It places value in close contests, which I like. Whether others want to place a value like that in the standings to determine the conference champion is the tough part. We all like close matches and want to be rewarded for such. Besides, everyone on here uses set scores and game scores to argue who is better anyways. Why not use it? Interesting. My math (which also needs checking) is a little different. I've tried to explain my calculations in red above. I have also compared this standings to the existing standings to show where teams have moved up or down, as follows: 1- Wisconsin (0) 2- Purdue (0) 3- Illinois (0) 3- Nebraska (+2) 5- Penn State (-2) 5- Michigan (+1) 7- Michigan State (-1) 7- Ohio State (-1) 9- Northwestern (0) 9- Minnesota (+2) 11- Indiana (-2) 12- Iowa (0 12- Maryland (0) 14- Rutgers (0) If I have done this correctly, it seems to indicate at this point in the season that the points system affects the top and bottom teams the least, and shuffles the middle teams somewhat. I am not a statistician, but I guess this makes sense if it is true that the better teams sweep the most and the worst teams get swept the most.
|
|
bluepenquin
Hall of Fame
4-Time VolleyTalk Poster of the Year (2019, 2018, 2017, 2016), All-VolleyTalk 1st Team (2021, 2020, 2019, 2018, 2017, 2016)
Posts: 12,440
|
Post by bluepenquin on Oct 24, 2014 11:09:12 GMT -5
Has there ever been discussion about any NCAA conferences adopting the international system used in tournaments and leagues to award different points in the conference standings depending on the outcome of the match, i.e., 3-0 or 3-1 = 3 points to winner, 0 to loser; 3-2 = 2 points to winner, 1 point to loser? It seems to me that over a full conference schedule such a system would reward and differentiate the stronger teams. Thoughts? As I understand this - use this to determine the conference champion? I don't think that is necessary. The Goal of volleyball is to win matches, not win a higher % of points, win more sets, or even be the team most likely to win if played on a neutral court. It should simply be the team that has the best record. If we want a system that picks the best team (can be different than the conference champion), then this may be an improvement - but we have other metrics that are better. But again - the goal is to win matches and record is the best way to do this (no matter the luck factor involved). I am not familar with the International setup - but if you are playing round robin matches to determine seeding for a tournament - then 3-2-1 makes sense. But that is not what we are doing with conferences championships.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 24, 2014 11:26:41 GMT -5
Has there ever been discussion about any NCAA conferences adopting the international system used in tournaments and leagues to award different points in the conference standings depending on the outcome of the match, i.e., 3-0 or 3-1 = 3 points to winner, 0 to loser; 3-2 = 2 points to winner, 1 point to loser? It seems to me that over a full conference schedule such a system would reward and differentiate the stronger teams. Thoughts? As I understand this - use this to determine the conference champion? I don't think that is necessary. The Goal of volleyball is to win matches, not win a higher % of points, win more sets, or even be the team most likely to win if played on a neutral court. It should simply be the team that has the best record. If we want a system that picks the best team (can be different than the conference champion), then this may be an improvement - but we have other metrics that are better. But again - the goal is to win matches and record is the best way to do this (no matter the luck factor involved). I am not familar with the International setup - but if you are playing round robin matches to determine seeding for a tournament - then 3-2-1 makes sense. But that is not what we are doing with conferences championships. I am not familiar with all of the leagues in other countries, but several, like Germany, Sweden, and France, use systems like this to calculate conference standings during regular league play. I guess what it comes down to is whether or not you think conceptually that the quality of wins and losses should affect conference standings (as opposed to just tournament seeding), i.e., a team which loses a close match should be rewarded, or a team that wins a close match should be penalized vs. a team that wins more easily. A point system is one method to measure the quality of wins and losses, but certainly not the only or possibly even the best way. This has been an interesting discussion.
|
|
|
Post by badgerbreath on Oct 24, 2014 12:12:01 GMT -5
Has there ever been discussion about any NCAA conferences adopting the international system used in tournaments and leagues to award different points in the conference standings depending on the outcome of the match, i.e., 3-0 or 3-1 = 3 points to winner, 0 to loser; 3-2 = 2 points to winner, 1 point to loser? It seems to me that over a full conference schedule such a system would reward and differentiate the stronger teams. Thoughts? As I understand this - use this to determine the conference champion? I don't think that is necessary. The Goal of volleyball is to win matches, not win a higher % of points, win more sets, or even be the team most likely to win if played on a neutral court. It should simply be the team that has the best record. If we want a system that picks the best team (can be different than the conference champion), then this may be an improvement - but we have other metrics that are better. But again - the goal is to win matches and record is the best way to do this (no matter the luck factor involved). I am not familiar with the International setup - but if you are playing round robin matches to determine seeding for a tournament - then 3-2-1 makes sense. But that is not what we are doing with conferences championships. I think it would make sense however to use a formula like this as a tie breaker between two teams with the same W-L record...and it is indeed very "American" (as C4ndelight put it) to want to avoid ties of any sort. Maybe not like the NCAA however. It only makes sense in a true round robin set up with balanced schedules home and away, IMO. It would certainly add a different dynamic to games. Suddenly just getting to five sets, or keeping a match at four or fewer sets, yields points. Set point at end of set four becomes an even bigger a game within a game. I wonder how that would change strategies? Or if it would.
|
|
|
Post by ja on Oct 24, 2014 12:25:51 GMT -5
This is America. Losers don't get hugs, and they certainly don't get a point either. But losers do get orange slices And meal money . On the serious note, I think this is how DR got rewarded this year for keep fighting! As a coach, you will have now tool to encourage you team and get reward for hard work. How many times one bad call in 5 set was the difference between W and L? With this system you can get at least one point. But I can put my money that this would not happened in any close future.
|
|
bluepenquin
Hall of Fame
4-Time VolleyTalk Poster of the Year (2019, 2018, 2017, 2016), All-VolleyTalk 1st Team (2021, 2020, 2019, 2018, 2017, 2016)
Posts: 12,440
|
Post by bluepenquin on Oct 24, 2014 12:44:44 GMT -5
As I understand this - use this to determine the conference champion? I don't think that is necessary. The Goal of volleyball is to win matches, not win a higher % of points, win more sets, or even be the team most likely to win if played on a neutral court. It should simply be the team that has the best record. If we want a system that picks the best team (can be different than the conference champion), then this may be an improvement - but we have other metrics that are better. But again - the goal is to win matches and record is the best way to do this (no matter the luck factor involved). I am not familar with the International setup - but if you are playing round robin matches to determine seeding for a tournament - then 3-2-1 makes sense. But that is not what we are doing with conferences championships. I am not familiar with all of the leagues in other countries, but several, like Germany, Sweden, and France, use systems like this to calculate conference standings during regular league play. I guess what it comes down to is whether or not you think conceptually that the quality of wins and losses should affect conference standings (as opposed to just tournament seeding), i.e., a team which loses a close match should be rewarded, or a team that wins a close match should be penalized vs. a team that wins more easily. A point system is one method to measure the quality of wins and losses, but certainly not the only or possibly even the best way. This has been an interesting discussion. You also have the situation of which match below was closer: 25-23, 25-23, 25-23 25-15, 23-25, 25-18, 23-25, 15-8 25-15, 23-25, 25-18, 25-23 I do think 5th sets are too much of a crapshot, but I guess if you don't want to lose a 5th set, just make sure you win 3 of the 1st 4.
|
|
|
Post by redbeard2008 on Oct 24, 2014 13:06:27 GMT -5
Could play matches to four 21-pt sets, with a tie-breaker 15-pt set,
4-0 - 5 points 3-1 - 4 points 3-2 - 3 points 2-3 - 2 points 1-3 - 1 point 0-4 - 0 points
|
|
|
Post by pancake83 on Oct 24, 2014 14:49:15 GMT -5
JVA is starting a Wisconsin Cup (new version of their power league) with this scoring system in place for advancing in their three total stops... Interesting concept for a Power league
|
|
|
Post by The Bofa on the Sofa on Oct 27, 2014 15:03:55 GMT -5
I have been out of town so haven't had a chance to engage this thread, but there are things I can say (of course).
The question of using this type of approach depends on what you want to determine. If all you want to do is to reflect match wins and losses, then that's what you use. However, if you are interested in determining who the best team is, something like this is a much better approach. I don't know if this breakdown in itself is the best, but it's likely better than simple wins and losses.
I have been meaning to do a post on that at some point recently, but it would be easier with figures. The question that we have to ask is, what does who won tell us about the teams? In fact, very little, this is clear. I know people don't like to hear that, and even in this thread, there are comments like "all that matters is who won." But the reason that is all that matters is because we base things like conference champions and tournament selections and who advances in the tournament on who won the match. However, there is no apriori reason why we have to do any of these, nor is there anything that says these are the best ways (although I think it's fair to argue that the team that wins the match in a tournament is the team to advance, since that is inherently what single elimination tournament means).
How do I know that wins is not all that matters in determining the better team? I have investigated it. As many of you know, I have a database of about 25000 "match pairs", which are when two teams play two matches in the same year. I have used this to investigate things like, how often the same team wins both matches, under a variety of different conditions. I can examine the effect of date, score and location.
For example, I can tell you that, overall, the team that wins the first match wins the next one 76.15% of the time. Now, this is distributed over all ranges of time (the average time between matches is 30 days), and all venues. Also, that's distorted a little bit due to home court issues (the team that wins first is more likely to be at home, where they have an advantage, and therefore the second match is more likely to be on the road, where they have a disadvantage), but I can narrow it down to match pairs where both are played on neutral courts. In this case, the second team wins 76.8% of the time, so you can see that, indeed, there is a little distortion in the overall. But it doesn't matter to the point.
So that is the outcome you get if you look to the team that won the first match. However, what if we look at the team that scored the most points in the first match, regardless of whether they win or lose? In that case, that team wins the second match 76.65% of the time. IOW, a team who outscores their opponent in the first match is more likely to win the second time they play than the team that won. It's not a lot, but it's real.
If we just look at neutral/neutral outcomes, it is the same. Where the winning team wins 76.8% of the time, the team that scores more wins 77.1% of the time when they play again. So points is a better better predictor of the future than wins.
The same is true for sets, to the point of this thread. The team that wins in three wins about 85% when they play again. However, the winner of a 5 set match only goes on to win 59% of the rematches (the team that scores more in a 5 set match goes on to win 62% of the time). Clearly, clearly, clearly there is a difference in the outcome of a 5 set match as opposed to three in terms of what it tells you about the teams. This is undeniable. Teams that go to 5 sets are more similar than when one team wins in three.
Ultimately, this is a reflection of the point differences, but lacking the points themselves, sets are a proxy. In the same way, lacking sets information, matches won are a proxy for determining who was better. Just not as good as more detailed information. However, there is a high degree of correlation between who won and who scored more points, and that's why it works as well as it does.
So to summarize, if you want the standings to better reflect which teams are better, some sort of assessment of sets/points will be better than just win/loss. However, if all you want to know is who won and lost, then use wins and losses.
|
|
|
Post by tomclen on Oct 27, 2014 15:58:59 GMT -5
I think it is safe to guess that the answer is "no". NCAA is not going to do this any time before the sun turns into a red giant. Isn't the sun already sort of a red giant?
|
|
|
Post by badgerbreath on Oct 27, 2014 16:04:16 GMT -5
No. It's a yellow dwarf, which is a disturbing image.
|
|
|
Post by mikegarrison on Oct 27, 2014 16:33:45 GMT -5
I think it is safe to guess that the answer is "no". NCAA is not going to do this any time before the sun turns into a red giant. Isn't the sun already sort of a red giant? Nope. But it will be. Right now it is fusing hydrogen and the radiation from that balances its gravity. When that slows down enough that it doesn't balance the gravity, it will collapse. Paradoxically, by collapsing it will heat up the core of the sun enough that that the outer shell of the sun will expand. But it will have a cooler surface temperature, so it will be a red giant instead of a main-branch yellow star. en.wikipedia.org/wiki/Stellar_evolution
|
|
|
Post by tomclen on Oct 27, 2014 16:49:43 GMT -5
Isn't the sun already sort of a red giant? Nope. But it will be. Right now it is fusing hydrogen and the radiation from that balances its gravity. When that slows down enough that it doesn't balance the gravity, it will collapse. Paradoxically, by collapsing it will heat up the core of the sun enough that that the outer shell of the sun will expand. But it will have a cooler surface temperature, so it will be a red giant instead of a main-branch yellow star. en.wikipedia.org/wiki/Stellar_evolutionSo, will we still have sand volleyball?
|
|
|
Post by mikegarrison on Oct 27, 2014 16:57:43 GMT -5
Nope. But it will be. Right now it is fusing hydrogen and the radiation from that balances its gravity. When that slows down enough that it doesn't balance the gravity, it will collapse. Paradoxically, by collapsing it will heat up the core of the sun enough that that the outer shell of the sun will expand. But it will have a cooler surface temperature, so it will be a red giant instead of a main-branch yellow star. en.wikipedia.org/wiki/Stellar_evolutionSo, will we still have sand volleyball? Probably more like glass volleyball at that point, since surface temperatures on the Earth will be something close to what Mercury is like today. (Mercury and Venus will be completely gone.)
|
|