|
Post by mikegarrison on Jan 3, 2017 18:12:34 GMT -5
There is no "cap" for SOS, as far as I am aware. Do you just mean that in practice the highest SOSs run to that range? There IS a SOS cap for teams playing in conferences. Not counting your own results against them, your conference opponents will have an equal number of wins and losses in conference (assuming balance schedule) or close to that if imbalanced schedule. Your SOS is .500 for conference matches (for balanced conference schedules) or close to that for unbalanced conference schedules. There is a mathematical "cap" to how much you can improve (or reduce that) that, but it depends on exactly how many non-conference matches you play. The numbers he mentioned are an accurate SOS range, with theoretical max caps slightly higher. Let's try a simple example. 10 teams in conference, playing a full round robin. Each team also has 10 non-conference matches. Team A (the one of interest) goes 18-0 in conference. The total number of wins and losses in the conference is 180 and 180. The other teams collectively go 162-180, but since Team A's 18 wins are removed, they go 162-162, so .500. If they all won all their OoC matches, that would add 90 wins, so 252-162. The ten OoC opponents of team A could all have perfect records (except for Team A) and be something like 270-0. So that's 522-162. So the theoretical practical limit for SOS is somewhere about .750; more if your conference plays an unbalanced schedule and you avoid multiple matches against the big losers. But what is holding you back is your conference opponents, not your OoC opponents. So in this case playing Creighton would have only been a negative, because they had six losses before the tournament and I specified that all the other OoC opponents had perfect records. Where I am going with this is that I don't think Creighton is a perfect RPI booster. For instance, American University had basically the same record as Creighton and yet would have been a less risky threat to play.
|
|
|
Post by c4ndlelight on Jan 3, 2017 18:17:40 GMT -5
I don't follow what you mean here. Basic RPI doesn't care if you win or lose a particular match. A loss is a loss and a win is a win. You get the benefit (or penalty) of your opponent's record in your SOS regardless of whether you win or lose. Do you mean that their expected record will benefit RPI through SOS more than a loss will hurt RPI? It seems like that obviously depends on the team in question. For instance, if team A has a perfect record and also an SOS percentage higher than Creighton's expected W-L %, then they would only lose RPI by playing them regardless of the outcome of the match. (Of course, such a team would already have an insanely high RPI.) The line of logic I had is:
1) losing a match to a team with a very good overall record (in this example Creighton for next year) will have a very small (negative) impact on a team's final RPI compared to not playing the match. 2) winning a match against a team with a very good overall record has a strong positive impact to the team's RPI compared to not playing the match. 3) There are other potential aspects - potential Top 25 win and the bonus points that go along with this. Also may make the difference in terms of scheduling bonus points (although there are many better options for this). 4) There are plenty of bad teams (bad records) where you are much worse off playing the match and winning compared to not playing the match at all.
If a top team has something like a 50% chance of beating Creighton - than this is a great risk/reward opportunity for the team. A 15-25% chance of winning probably makes this a good decision to schedule (a Top 15 team is probably going to be something close to 50%). Given the conference for Creighton - there is probably a very low chance that Creighton doesn't have a really good record, they may be one of the better bets for a very good record and for being a Top 25 opponent.
Teams to avoid for RPI are teams from the Big Ten or Pac 12. There is the risk that they will have a middling record despite being a quality opponent - and if they do have a really good record than they become very hard to beat. Creighton may be difficult to beat, but they are not likely to be elite level - and seems like a pretty good bet to have a very good record. Hard to find many other Top 25 teams that are both 'beatable' and have a high % chance of being a top 25 team. For this reason - any and all top teams should want Creighton on their schedule (Creighton needs to be 'choosey' on who they want to play, they should have the power). You can probably throw Kansas, North Carolina, and Kentucky into a similar boat - everyone should be trying to schedule these teams, so these teams should have the power in determining who they get to play. USC is benefited more by playing Creighton and Kentucky than Creighton and Kentucky are benefited by playing USC. Way too much risk associated with playing a team like USC - they can be elite or very hard to beat and they can also end up with a middling record because of the hard conference they play in.
The benefit to Creighton/UNI in playing USC is not in RPI, but prestige. It's valuable for KBB to tell recruits - the Big East may not be great, but we play top PAC and SEC teams every year. There's also the soft metrics - a team from a lesser conference needs to beat a team from a top conference to get that seed.
|
|
|
Post by c4ndlelight on Jan 3, 2017 18:19:21 GMT -5
There IS a SOS cap for teams playing in conferences. Not counting your own results against them, your conference opponents will have an equal number of wins and losses in conference (assuming balance schedule) or close to that if imbalanced schedule. Your SOS is .500 for conference matches (for balanced conference schedules) or close to that for unbalanced conference schedules. There is a mathematical "cap" to how much you can improve (or reduce that) that, but it depends on exactly how many non-conference matches you play. The numbers he mentioned are an accurate SOS range, with theoretical max caps slightly higher. Let's try a simple example. 10 teams in conference, playing a full round robin. Each team also has 10 non-conference matches. Team A (the one of interest) goes 18-0 in conference. The total number of wins and losses in the conference is 180 and 180. The other teams collectively go 162-180, but since Team A's 18 wins are removed, they go 162-162, so .500. If they all won all their OoC matches, that would add 90 wins, so 252-162. The ten OoC opponents of team A could all have perfect records (except for Team A) and be something like 270-0. So that's 522-162. So the theoretical practical limit for SOS is somewhere about .750; more if your conference plays an unbalanced schedule and you avoid multiple matches against the big losers. But what is holding you back is your conference opponents, not your OoC opponents. So in this case playing Creighton would have only been a negative, because they had six losses before the tournament and I specified that all the other OoC opponents had perfect records. Where I am going with this is that I don't think Creighton is a perfect RPI booster. For instance, American University had basically the same record as Creighton and yet would have been a less risky threat to play. You have to take into account: 1) RPI bonus points. Beating a Top 25 opponent like Creighton would net about .028 points (I think that's it), so that needs to be put into the math. 2) Additional impact on SOS - in addition to being factored in mathematically, the Committee looks at your raw SOS and OOC SOS numbers to make seed-line and at-large decisions* (*unless you're Penn St.), plus the value in playing a Top 25 match on your Nitty Gritty. American, being on the wrong end of the Top 100, doesn't do much beyond the mathematical inputs. That said, I agree American U is a great choice to play. The problem is, there aren't a lot of teams that are easy to beat that you can rely on having that great of a record. Not everyone can play American and similar options are likely to have unexpected down years. If Creighton or USC or Kentucky has a down year in 2017, they'll still be Top 50. UNI won't be that far outside of it.
|
|
|
Post by mikegarrison on Jan 3, 2017 18:33:43 GMT -5
Yeah, it was clear that this year the committee made several choices that went away from straight basic RPI and rewarded teams for playing an OoC schedule that the committee liked better.
Personally, that bothered me a lot. Not because I think RPI is so great, but because when the committee has the leeway to choose multiple factors without transparency or restriction by precedence, it leaves things wide open to seeding the tournament based on name recognition and committee bias. It means that when the teams the committee wants to seed have high RPIs, the committee can just seed them, and when the teams the committee wants to seed have lower RPIs, the committee can seed them anyway. So what are you supposed to do it you aren't Penn State, Texas, Stanford, or Florida?
Even worse, if you *are* Penn State, Texas, Stanford, or Florida you can arrange to play among yourselves in the OoC and then take mutual credit (and get rewarded) for your tough OoC scheduling. This inbred scheduling is what RPI was invented to stop (in men's basketball).
|
|
|
Post by azvolleydad on Jan 3, 2017 18:53:39 GMT -5
Wow -- Washington is practically untouched. Losing less than 2% of any significant category.
|
|
|
Post by mikegarrison on Jan 3, 2017 18:57:23 GMT -5
Wow -- Washington is practically untouched. Losing less than 2% of any significant category. They only had one senior. She was a transfer that came in this year, or else they would have had no seniors at all. And she rarely got into any matches. (FWIW they also had only one sophomore. So two years from now....)
|
|
|
Post by bbk on Jan 3, 2017 19:28:14 GMT -5
Let's try a simple example. 10 teams in conference, playing a full round robin. Each team also has 10 non-conference matches. Team A (the one of interest) goes 18-0 in conference. The total number of wins and losses in the conference is 180 and 180. The other teams collectively go 162-180, but since Team A's 18 wins are removed, they go 162-162, so .500. If they all won all their OoC matches, that would add 90 wins, so 252-162. The ten OoC opponents of team A could all have perfect records (except for Team A) and be something like 270-0. So that's 522-162. So the theoretical practical limit for SOS is somewhere about .750; more if your conference plays an unbalanced schedule and you avoid multiple matches against the big losers. But what is holding you back is your conference opponents, not your OoC opponents. So in this case playing Creighton would have only been a negative, because they had six losses before the tournament and I specified that all the other OoC opponents had perfect records. Where I am going with this is that I don't think Creighton is a perfect RPI booster. For instance, American University had basically the same record as Creighton and yet would have been a less risky threat to play. You have to take into account: 1) RPI bonus points. Beating a Top 25 opponent like Creighton would net about .028 points (I think that's it), so that needs to be put into the math. 2) Additional impact on SOS - in addition to being factored in mathematically, the Committee looks at your raw SOS and OOC SOS numbers to make seed-line and at-large decisions* (*unless you're Penn St.), plus the value in playing a Top 25 match on your Nitty Gritty. American, being on the wrong end of the Top 100, doesn't do much beyond the mathematical inputs. That said, I agree American U is a great choice to play. The problem is, there aren't a lot of teams that are easy to beat that you can rely on having that great of a record. Not everyone can play American and similar options are likely to have unexpected down years. If Creighton or USC or Kentucky has a down year in 2017, they'll still be Top 50. UNI won't be that far outside of it. So then a school some may try to schedule as well could be Pepperdine? BTW this ia a masterclass in how to schedule. Many thanks
|
|
|
Post by c4ndlelight on Jan 3, 2017 20:02:39 GMT -5
You have to take into account: 1) RPI bonus points. Beating a Top 25 opponent like Creighton would net about .028 points (I think that's it), so that needs to be put into the math. 2) Additional impact on SOS - in addition to being factored in mathematically, the Committee looks at your raw SOS and OOC SOS numbers to make seed-line and at-large decisions* (*unless you're Penn St.), plus the value in playing a Top 25 match on your Nitty Gritty. American, being on the wrong end of the Top 100, doesn't do much beyond the mathematical inputs. That said, I agree American U is a great choice to play. The problem is, there aren't a lot of teams that are easy to beat that you can rely on having that great of a record. Not everyone can play American and similar options are likely to have unexpected down years. If Creighton or USC or Kentucky has a down year in 2017, they'll still be Top 50. UNI won't be that far outside of it. So then a school some may try to schedule as well could be Pepperdine? BTW this ia a masterclass in how to schedule. Many thanks Pepperdine would be a bad choice. 13-17 this year, even if they improve I doubt they'll get past 20 wins. In that conference, BYU/San Diego are solid options. They're also tough to beat though.
|
|
|
Post by trollhunter on Jan 3, 2017 20:10:35 GMT -5
You have to take into account: 1) RPI bonus points. Beating a Top 25 opponent like Creighton would net about .028 points (I think that's it), so that needs to be put into the math. 2) Additional impact on SOS - in addition to being factored in mathematically, the Committee looks at your raw SOS and OOC SOS numbers to make seed-line and at-large decisions* (*unless you're Penn St.), plus the value in playing a Top 25 match on your Nitty Gritty. American, being on the wrong end of the Top 100, doesn't do much beyond the mathematical inputs. That said, I agree American U is a great choice to play. The problem is, there aren't a lot of teams that are easy to beat that you can rely on having that great of a record. Not everyone can play American and similar options are likely to have unexpected down years. If Creighton or USC or Kentucky has a down year in 2017, they'll still be Top 50. UNI won't be that far outside of it. So then a school some may try to schedule as well could be Pepperdine? BTW this ia a masterclass in how to schedule. Many thanks No, Pepperdine is typically not a good team to schedule. While they may flirt with RPI top 50, it is not consistent. And the most important thing for RPI scheduling is to schedule teams that will usually have a great W/L record (and you have a decent change to beat). Creighton falls into this "good/great" to schedule for RPI. There may be a few better teams to schedule, but not many. Also, as MikeG pointed out - it is really imperative for your entire conference, from top to bottom, schedule for wins in OOC play. That is something that B1G and SEC have done well the past few years, with a couple schools as notable exceptions. Some conferences discuss this as a group in conference meetings, others conferences don't and get a couple conference teams with 3-25 records (0-12 in non-conference) dragging everyone down because they must play them twice.
|
|
|
Post by c4ndlelight on Jan 3, 2017 20:17:56 GMT -5
So then a school some may try to schedule as well could be Pepperdine? BTW this ia a masterclass in how to schedule. Many thanks No, Pepperdine is typically not a good team to schedule. While they may flirt with RPI top 50, it is not consistent. And the most important thing for RPI scheduling is to schedule teams that will usually have a great W/L record (and you have a decent change to beat). Creighton falls into this "good/great" to schedule for RPI. There may be a few better teams to schedule, but not many. Also, as MikeG pointed out - it is really imperative for your entire conference, from top to bottom, schedule for wins in OOC play. That is something that B1G and SEC have done well the past few years, with a couple schools as notable exceptions. Some conferences discuss this as a group in conference meetings, others conferences don't and get a couple conference teams with 3-25 records dragging everyone down because they must play them twice. The PAC-12 has gotten much better at it since 2010 (the year the entire conference got smacked by the Committee with Top 10-type Washington and UCLA teams didn't get seeded and a Top 25 Oregon getting left out entirely). Colorado has been the biggest exception, but that might improve under the new staff. Oregon St. has overreached at times, but the conference got in line.
|
|
|
Post by trollhunter on Jan 3, 2017 20:22:04 GMT -5
No, Pepperdine is typically not a good team to schedule. While they may flirt with RPI top 50, it is not consistent. And the most important thing for RPI scheduling is to schedule teams that will usually have a great W/L record (and you have a decent change to beat). Creighton falls into this "good/great" to schedule for RPI. There may be a few better teams to schedule, but not many. Also, as MikeG pointed out - it is really imperative for your entire conference, from top to bottom, schedule for wins in OOC play. That is something that B1G and SEC have done well the past few years, with a couple schools as notable exceptions. Some conferences discuss this as a group in conference meetings, others conferences don't and get a couple conference teams with 3-25 records dragging everyone down because they must play them twice. The PAC-12 has gotten much better at it since 2010 (the year the entire conference got smacked by the Committee with Top 10-type Washington and UCLA teams didn't get seeded and a Top 25 Oregon getting left out entirely). Colorado has been the biggest exception, but that might improve under the new staff. Oregon St. has overreached at times, but the conference got in line. Yes you are correct I should have added PAC to the list. ACC actually scheduled pretty well this year too, not sure if it is a trend or an anomaly.
|
|
|
Post by Northern lights on Jan 3, 2017 20:28:12 GMT -5
I think Pitt is going to be a good team next season. They have young talent, and Dan Fisher is a great Head Coach. He is all in with Volleyball. I think they are top 25 next year.
|
|
|
Post by Fight On! on Jan 3, 2017 20:31:14 GMT -5
I don't follow what you mean here. Basic RPI doesn't care if you win or lose a particular match. A loss is a loss and a win is a win. You get the benefit (or penalty) of your opponent's record in your SOS regardless of whether you win or lose. Do you mean that their expected record will benefit RPI through SOS more than a loss will hurt RPI? It seems like that obviously depends on the team in question. For instance, if team A has a perfect record and also an SOS percentage higher than Creighton's expected W-L %, then they would only lose RPI by playing them regardless of the outcome of the match. (Of course, such a team would already have an insanely high RPI.) The line of logic I had is:
1) losing a match to a team with a very good overall record (in this example Creighton for next year) will have a very small (negative) impact on a team's final RPI compared to not playing the match. 2) winning a match against a team with a very good overall record has a strong positive impact to the team's RPI compared to not playing the match. 3) There are other potential aspects - potential Top 25 win and the bonus points that go along with this. Also may make the difference in terms of scheduling bonus points (although there are many better options for this). 4) There are plenty of bad teams (bad records) where you are much worse off playing the match and winning compared to not playing the match at all.
If a top team has something like a 50% chance of beating Creighton - than this is a great risk/reward opportunity for the team. A 15-25% chance of winning probably makes this a good decision to schedule (a Top 15 team is probably going to be something close to 50%). Given the conference for Creighton - there is probably a very low chance that Creighton doesn't have a really good record, they may be one of the better bets for a very good record and for being a Top 25 opponent.
Teams to avoid for RPI are teams from the Big Ten or Pac 12. There is the risk that they will have a middling record despite being a quality opponent - and if they do have a really good record than they become very hard to beat. Creighton may be difficult to beat, but they are not likely to be elite level - and seems like a pretty good bet to have a very good record. Hard to find many other Top 25 teams that are both 'beatable' and have a high % chance of being a top 25 team. For this reason - any and all top teams should want Creighton on their schedule (Creighton needs to be 'choosey' on who they want to play, they should have the power). You can probably throw Kansas, North Carolina, and Kentucky into a similar boat - everyone should be trying to schedule these teams, so these teams should have the power in determining who they get to play. USC is benefited more by playing Creighton and Kentucky than Creighton and Kentucky are benefited by playing USC. Way too much risk associated with playing a team like USC - they can be elite or very hard to beat and they can also end up with a middling record because of the hard conference they play in.
So Mick is a scheduling wizard??
|
|
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: 12,447
|
Post by bluepenquin on Jan 3, 2017 22:18:03 GMT -5
The line of logic I had is:
1) losing a match to a team with a very good overall record (in this example Creighton for next year) will have a very small (negative) impact on a team's final RPI compared to not playing the match. 2) winning a match against a team with a very good overall record has a strong positive impact to the team's RPI compared to not playing the match. 3) There are other potential aspects - potential Top 25 win and the bonus points that go along with this. Also may make the difference in terms of scheduling bonus points (although there are many better options for this). 4) There are plenty of bad teams (bad records) where you are much worse off playing the match and winning compared to not playing the match at all.
If a top team has something like a 50% chance of beating Creighton - than this is a great risk/reward opportunity for the team. A 15-25% chance of winning probably makes this a good decision to schedule (a Top 15 team is probably going to be something close to 50%). Given the conference for Creighton - there is probably a very low chance that Creighton doesn't have a really good record, they may be one of the better bets for a very good record and for being a Top 25 opponent.
Teams to avoid for RPI are teams from the Big Ten or Pac 12. There is the risk that they will have a middling record despite being a quality opponent - and if they do have a really good record than they become very hard to beat. Creighton may be difficult to beat, but they are not likely to be elite level - and seems like a pretty good bet to have a very good record. Hard to find many other Top 25 teams that are both 'beatable' and have a high % chance of being a top 25 team. For this reason - any and all top teams should want Creighton on their schedule (Creighton needs to be 'choosey' on who they want to play, they should have the power). You can probably throw Kansas, North Carolina, and Kentucky into a similar boat - everyone should be trying to schedule these teams, so these teams should have the power in determining who they get to play. USC is benefited more by playing Creighton and Kentucky than Creighton and Kentucky are benefited by playing USC. Way too much risk associated with playing a team like USC - they can be elite or very hard to beat and they can also end up with a middling record because of the hard conference they play in.
The benefit to Creighton/UNI in playing USC is not in RPI, but prestige. It's valuable for KBB to tell recruits - the Big East may not be great, but we play top PAC and SEC teams every year. There's also the soft metrics - a team from a lesser conference needs to beat a team from a top conference to get that seed. Sure, there are multiple reasons other than RPI for a team to schedule. Maybe the most important is scheduling such that best prepares your team for the rest of the season. In theory, if you have a younger team - one may want to be able to try different rotations and may not want to play the best teams to start the year? I was just speaking in terms of maximizing RPI. And Creighton is in a tough situation because the Big East could be pretty bad next year and they will have no opportunities for Top 25 and 50 wins in conference - and they will have to have some in order to get a seed. That said - they need to maximize their RPI and need to be very careful on where they get their Top 25 wins. Teams like Stanford, Texas and some others are probably too good to mess with and I think USC is too unpredictable. Agree that they probably need to play a couple teams from the B1G or Pac, but I would limit the number of those matches.
|
|
|
Post by n00b on Jan 3, 2017 22:31:11 GMT -5
The benefit to Creighton/UNI in playing USC is not in RPI, but prestige. It's valuable for KBB to tell recruits - the Big East may not be great, but we play top PAC and SEC teams every year. There's also the soft metrics - a team from a lesser conference needs to beat a team from a top conference to get that seed. Sure, there are multiple reasons other than RPI for a team to schedule. Maybe the most important is scheduling such that best prepares your team for the rest of the season. In theory, if you have a younger team - one may want to be able to try different rotations and may not want to play the best teams to start the year? I was just speaking in terms of maximizing RPI. And Creighton is in a tough situation because the Big East could be pretty bad next year and they will have no opportunities for Top 25 and 50 wins in conference - and they will have to have some in order to get a seed. That said - they need to maximize their RPI and need to be very careful on where they get their Top 25 wins. Teams like Stanford, Texas and some others are probably too good to mess with and I think USC is too unpredictable. Agree that they probably need to play a couple teams from the B1G or Pac, but I would limit the number of those matches. The Big East could easily be a one-big league next year. Marquette loses Louis and Xavier loses two of their best attackers. Villanova and Butler should be improved, but I'm not sure if they'll get to at large caliber. Did Hawaii ever get a seed in a year that their conference was a one-bid league?
|
|