OrangeXtreme
The Mayor of Dewitt
- Joined
- Aug 15, 2011
- Messages
- 227,904
- Like
- 408,780
NCAAT Expansion
- Thread starter Cappy3
- Start date
OrangeXtreme
The Mayor of Dewitt
- Joined
- Aug 15, 2011
- Messages
- 227,904
- Like
- 408,780
Orangezoo
In the wind
- Joined
- Aug 26, 2011
- Messages
- 41,276
- Like
- 96,799
That’s cool - put everything in one place very interesting that includes the NIT….
cusefan88
All American
- Joined
- Feb 20, 2012
- Messages
- 6,718
- Like
- 17,915
Torvik is a great site. Exciting to see them get proper analytics involved.
cusefan88
All American
- Joined
- Feb 20, 2012
- Messages
- 6,718
- Like
- 17,915
Copied this from an article:
Wins Above Bubble is not all that different than the idea of Wins Above Replacement in baseball. It’s the measurement of how much better you are than a team that would be right on the cut line. How many more wins do you have than a team that would be on the edge of missing the tournament.
OrangeXtreme
The Mayor of Dewitt
- Joined
- Aug 15, 2011
- Messages
- 227,904
- Like
- 408,780
jncuse
I brought the Cocaine to the White House
- Joined
- Feb 19, 2012
- Messages
- 19,918
- Like
- 34,185
TLDR Summary based on my look at comparative rankings shown below
1. Barttorvik will do much the same as NET/KP - heavily impacted by margin. Wouldn't have been popular here last year.
2. People should like WAB WAB is calculated very differently than NET/KP/BART - more based on your actual W/L record and not really impacted by margin. WAB treated us better last year
In theory I think WAB should be a better measure in comparing P6 teams as it is more related to actual W/L records and schedule levels while different are in the same stratosphere.. I don't think it would be that strong to rank poorer conference teams or to rank 360 teams - you need margin to try to compare those teams.
---------------------------------------------------------------------------------
A quick look at the two new rankings. I think people should like the WAB based on the complaints of the NET. The NET criticizers won't like BARTTORVIK.
Barttorvik will be much of same old, based on Margin. Wins Above Bubble will be a very different ranking that is much more based on pure W/L.
#1. Barttorvik is heavily based on margin - its not going to produce significantly different rankings than KP Or NET. It was not friendly to Syracuse last year just like NET and KP. Barttorvik will have the same strengths and weaknesses as those two ranking systems
#2 Wins Above Bubble doesn't seem to be heavily based on margin. It seems to be much more based on W/L record, so it was friendlier to Syracuse last year than Bart/KP/NET. There is a certain profile of team that this ranking system is much more friendlier to and we can tie it back to the good old KP luck. So that is a positive or at least a complement to others - now I don't have a great feel how it is calculated, but based on the data presented below that its not driven by margin at all. Its probably a strong ranking to compare P6 level teams who have stronger schedules so you can run some sort of meaninful algorithm. Its probably pretty sketchy in ranking teams from mid to poor conferences, because its almost impossible to compare those type of teams to power teams without using margins
#3. Both rankings are still highly impacted by OOC play by your conference as a whole.
Some data to support my analysis above.
Barttorvik/NET/KP/Wins Above Bubble / KP Luck
Syracuse - 71/84/80/53/7
Utah St - 58/38/51/17/8
South Carolina - 50/51/54/11/4
Michigan St - 18/24/16/56/309
Just a reminder that "LUCK" is just an output derived from your expected W/L record based on your margin vs your actual W/L record.
Michigan St had great margins in their wins last year, with generally tight losses, so they were rewarded by NET/KP/Bart. They also had "bad" luck. The WAB doesn't reward them for their good margin - their BART was 18 but their WAB was 56.
Syracuse/Utah St/South Carolina did not have great margins last year - theirr ecords were much better than their margin suggests. The WAB doesn't punish them for their poor margin - the average BART for these 3 teams was 60 -- the average WAB was 27.
1. Barttorvik will do much the same as NET/KP - heavily impacted by margin. Wouldn't have been popular here last year.
2. People should like WAB WAB is calculated very differently than NET/KP/BART - more based on your actual W/L record and not really impacted by margin. WAB treated us better last year
In theory I think WAB should be a better measure in comparing P6 teams as it is more related to actual W/L records and schedule levels while different are in the same stratosphere.. I don't think it would be that strong to rank poorer conference teams or to rank 360 teams - you need margin to try to compare those teams.
---------------------------------------------------------------------------------
A quick look at the two new rankings. I think people should like the WAB based on the complaints of the NET. The NET criticizers won't like BARTTORVIK.
Barttorvik will be much of same old, based on Margin. Wins Above Bubble will be a very different ranking that is much more based on pure W/L.
#1. Barttorvik is heavily based on margin - its not going to produce significantly different rankings than KP Or NET. It was not friendly to Syracuse last year just like NET and KP. Barttorvik will have the same strengths and weaknesses as those two ranking systems
#2 Wins Above Bubble doesn't seem to be heavily based on margin. It seems to be much more based on W/L record, so it was friendlier to Syracuse last year than Bart/KP/NET. There is a certain profile of team that this ranking system is much more friendlier to and we can tie it back to the good old KP luck. So that is a positive or at least a complement to others - now I don't have a great feel how it is calculated, but based on the data presented below that its not driven by margin at all. Its probably a strong ranking to compare P6 level teams who have stronger schedules so you can run some sort of meaninful algorithm. Its probably pretty sketchy in ranking teams from mid to poor conferences, because its almost impossible to compare those type of teams to power teams without using margins
#3. Both rankings are still highly impacted by OOC play by your conference as a whole.
Some data to support my analysis above.
Barttorvik/NET/KP/Wins Above Bubble / KP Luck
Syracuse - 71/84/80/53/7
Utah St - 58/38/51/17/8
South Carolina - 50/51/54/11/4
Michigan St - 18/24/16/56/309
Just a reminder that "LUCK" is just an output derived from your expected W/L record based on your margin vs your actual W/L record.
Michigan St had great margins in their wins last year, with generally tight losses, so they were rewarded by NET/KP/Bart. They also had "bad" luck. The WAB doesn't reward them for their good margin - their BART was 18 but their WAB was 56.
Syracuse/Utah St/South Carolina did not have great margins last year - theirr ecords were much better than their margin suggests. The WAB doesn't punish them for their poor margin - the average BART for these 3 teams was 60 -- the average WAB was 27.
Last edited:
jncuse
I brought the Cocaine to the White House
- Joined
- Feb 19, 2012
- Messages
- 19,918
- Like
- 34,185
I have no clue how it is calculated exactly, but based on comparing teams with large gaps between their BART (which is about the same as NET) and their WAB, "wins above bubbles" doesn't seem to cater to margin that much. Its more based on your W/L record.
I think people should like it more here based on the complaints last year about the NET. At a top level glance I think its a better measure for teams from power conferences. I don't think its a great measure for mid to low conference schools but they aren't typically fighting for the bubble anyway.
Last edited:
jncuse
I brought the Cocaine to the White House
- Joined
- Feb 19, 2012
- Messages
- 19,918
- Like
- 34,185
I got some long posts above, so I will isolate the fact here:
Syracuse BART/KP/NET in 2024 = 71/80/84
Syracuse WAB in 2024 = 53
Margin matters much less under WAB - that becomes evident when you look at the similar numbers for teams like South Carolina and Utah St who also didn't have great margin profiles but good relative W/L reocrd. They were rewarded in WAB. A team like Michigan St who had good margins overall but not a great W/L record. Accordingly their WAB was much worse.
Syracuse BART/KP/NET in 2024 = 71/80/84
Syracuse WAB in 2024 = 53
Margin matters much less under WAB - that becomes evident when you look at the similar numbers for teams like South Carolina and Utah St who also didn't have great margin profiles but good relative W/L reocrd. They were rewarded in WAB. A team like Michigan St who had good margins overall but not a great W/L record. Accordingly their WAB was much worse.