Post by Commissioner on Dec 15, 2013 23:16:53 GMT -5
Time to go back to a favorite peeve, which was discussed previously in this thread udtitanbasketball.freeforums.net/thread/207/out-conference-schedule-strength but I thought needed one with a more descriptive subject heading: How does RPI work, and what is the effect of Strength of Schedule on RPI?
What sparks this is a post on the Valpo Board by poster "Class of 2014":
www.valpofanzone.com/forum/index.php?topic=1608.msg39708#msg39708
"Class of 2014" doesn't understand the way RPI works.
Here is the basic RPI calculation: 25% W-L Pct + 50% Opponents' W-L Pct. + 25% Opponents' Opponents' W-L Pct. = RPI.
It's a bit more complex because road wins count as 1.4 wins for determining your own W-L Pct. (but not the W-L Pct for your opponents or opponents' opponents); road losses count 0.6 losses (again, only for your own W-L record); home wins count .6 and home losses 1.4 (again, only for your own W-L Pct., not for opponents). Additionally, games you play against your opponents are backed out of their W-L record. Also, games against non-DI teams don't count.
Conference RPI is calculated simply by adding up all the individual team RPI and dividing by the # of teams in the league.
From this, we see that Class of 2014 basically has the effect of Oakland's schedule backwards. Oakland's schedule boosts its own RPI, because the biggest part of RPI is opponents' W-L Pct., and that for Oakland is a very high number. However, once Oakland starts playing its league schedule, its poor W-L will hurt. Currently Oakland is 1-9 vs. D-I teams. Let's say they win their final 3 non-con games (they actually have a 4th game against D-II Defiance, but that won't count for RPI purposes so we can ignore it), to finish 4-9. Let's say Oakland goes 8-8 in conference for an overall D-I mark of 12-17. That means every time they play a Horizon team, their W-L Pct. that figure into year end RPI would be would be .414. Now, let's suppose that instead of Indiana, Michigan State, UNC, UCLA and Gonzaga, they had scheduled and beaten Savannah State, Bethune Cookman, Stetson, Grambling, and Marist. Their W-L record that would factor into every other HL team's RPI (twice!) would be 17-12 for a W-L Pct. of .586. That is the biggest part of RPI - 50%. (Note that the difference would be partially offset, because each Horizon league team will also factor in as 25% of RPI the W-L record of Oakland's opponents.) Still, it is likely that Oakland's tough non-con schedule will pull down the RPI of every other Horizon team this year, by lowering the W-L record of each Horizon team's opponents - they largest component of RPI.
But note that while Oakland pulls down other RPIs, it gets a boost in RPI. Because opponents' W-L pct. is twice as important to RPI as your own W-L Pct., you can often move up in RPI by playing and losing badly to a really good team, or drop in RPI by playing and beating a weak sister. Especially when you play on the road, because the road loss only counts as .6 of a loss for your own W-L Pct.
So what does all this do for Conference RPI? At this point, Oakland has had almost no effect on any other Horizon team's RPI because it hasn't played any other Horizon teams. (It might have some small effect from affecting the opponents' opponents' W-L Pct.) So at this point Oakland's tough schedule may be beneficial to the league - to the extent it has boosted Oakland's RPI (although we have no idea what Oakland's RPI would be if it had dropped, say, three of its killer opponents to play others - which others? road or home? would Oakland win or lose those games?) But it is impossible to know if, at the end of the year, Oakland's schedule will help or hurt league RPI. Will Oakland's stronger RPI offset the drag its poor W-L record puts on the other teams in the league? (Note that it is only a relatively small part of each team's RPI).
Anyway, we can figure out if Oakland is hurting or helping league RPI right now. We can't know if it would help it or hurt it if it didn't play the killer schedule, because we can't know who they would have scheduled instead, and how they would have done. But we can see Oakland's actual effect.
The post I quote above from the Valpo board came up in a discussion noting the league's RPI was 12th, just behind the MVC. The posters - not only "Class of 2014" - gave much credit to Oakland for boosting the Horizon's strength of schedule. Do they deserve it? No.
The Horizon's current conference RPI is .5096. Oakland's is .5117. So Oakland is slightly raising the conference RPI. Without Oakland, the other 8 teams would have a conference RPI of .5093. Where would that place the league? 12th - exactly where it is now.
In summary, Oakland has marginally improved the league's RPI. But the difference is so slight it would not change the league's conference RPI rank. If Oakland had instead played Milwaukee's non-con schedule and compiled the same record as Milwaukee, it would have improved the league's RPI much more. Oakland's tough schedule may have other benefits for the Horizon - TV exposure mainly. Those things can be discussed. But it is not raising in any meaningful way, if at all, the conference's overall RPI situation.
What sparks this is a post on the Valpo Board by poster "Class of 2014":
Oakland scedule
HL - Plus
OU - Negative (shoulda, coulda, woulda tonight)
I guess you can say Oakland has been the sacrificial lamb this year to help our league strength of schedule.
HL - Plus
OU - Negative (shoulda, coulda, woulda tonight)
I guess you can say Oakland has been the sacrificial lamb this year to help our league strength of schedule.
"Class of 2014" doesn't understand the way RPI works.
Here is the basic RPI calculation: 25% W-L Pct + 50% Opponents' W-L Pct. + 25% Opponents' Opponents' W-L Pct. = RPI.
It's a bit more complex because road wins count as 1.4 wins for determining your own W-L Pct. (but not the W-L Pct for your opponents or opponents' opponents); road losses count 0.6 losses (again, only for your own W-L record); home wins count .6 and home losses 1.4 (again, only for your own W-L Pct., not for opponents). Additionally, games you play against your opponents are backed out of their W-L record. Also, games against non-DI teams don't count.
Conference RPI is calculated simply by adding up all the individual team RPI and dividing by the # of teams in the league.
From this, we see that Class of 2014 basically has the effect of Oakland's schedule backwards. Oakland's schedule boosts its own RPI, because the biggest part of RPI is opponents' W-L Pct., and that for Oakland is a very high number. However, once Oakland starts playing its league schedule, its poor W-L will hurt. Currently Oakland is 1-9 vs. D-I teams. Let's say they win their final 3 non-con games (they actually have a 4th game against D-II Defiance, but that won't count for RPI purposes so we can ignore it), to finish 4-9. Let's say Oakland goes 8-8 in conference for an overall D-I mark of 12-17. That means every time they play a Horizon team, their W-L Pct. that figure into year end RPI would be would be .414. Now, let's suppose that instead of Indiana, Michigan State, UNC, UCLA and Gonzaga, they had scheduled and beaten Savannah State, Bethune Cookman, Stetson, Grambling, and Marist. Their W-L record that would factor into every other HL team's RPI (twice!) would be 17-12 for a W-L Pct. of .586. That is the biggest part of RPI - 50%. (Note that the difference would be partially offset, because each Horizon league team will also factor in as 25% of RPI the W-L record of Oakland's opponents.) Still, it is likely that Oakland's tough non-con schedule will pull down the RPI of every other Horizon team this year, by lowering the W-L record of each Horizon team's opponents - they largest component of RPI.
But note that while Oakland pulls down other RPIs, it gets a boost in RPI. Because opponents' W-L pct. is twice as important to RPI as your own W-L Pct., you can often move up in RPI by playing and losing badly to a really good team, or drop in RPI by playing and beating a weak sister. Especially when you play on the road, because the road loss only counts as .6 of a loss for your own W-L Pct.
So what does all this do for Conference RPI? At this point, Oakland has had almost no effect on any other Horizon team's RPI because it hasn't played any other Horizon teams. (It might have some small effect from affecting the opponents' opponents' W-L Pct.) So at this point Oakland's tough schedule may be beneficial to the league - to the extent it has boosted Oakland's RPI (although we have no idea what Oakland's RPI would be if it had dropped, say, three of its killer opponents to play others - which others? road or home? would Oakland win or lose those games?) But it is impossible to know if, at the end of the year, Oakland's schedule will help or hurt league RPI. Will Oakland's stronger RPI offset the drag its poor W-L record puts on the other teams in the league? (Note that it is only a relatively small part of each team's RPI).
Anyway, we can figure out if Oakland is hurting or helping league RPI right now. We can't know if it would help it or hurt it if it didn't play the killer schedule, because we can't know who they would have scheduled instead, and how they would have done. But we can see Oakland's actual effect.
The post I quote above from the Valpo board came up in a discussion noting the league's RPI was 12th, just behind the MVC. The posters - not only "Class of 2014" - gave much credit to Oakland for boosting the Horizon's strength of schedule. Do they deserve it? No.
The Horizon's current conference RPI is .5096. Oakland's is .5117. So Oakland is slightly raising the conference RPI. Without Oakland, the other 8 teams would have a conference RPI of .5093. Where would that place the league? 12th - exactly where it is now.
In summary, Oakland has marginally improved the league's RPI. But the difference is so slight it would not change the league's conference RPI rank. If Oakland had instead played Milwaukee's non-con schedule and compiled the same record as Milwaukee, it would have improved the league's RPI much more. Oakland's tough schedule may have other benefits for the Horizon - TV exposure mainly. Those things can be discussed. But it is not raising in any meaningful way, if at all, the conference's overall RPI situation.