So I created a spreadsheet to track game predictions based on the
Sagarin rating system. It's a pretty simple rating system for predictions. Take the home team's rating, subtract the visiting team's rating, add the home court advantage adjustment [he currently uses 3.10], and the result is basically his prediction of the home team's point spread.
Just wanted to be able to compare that quickly and easily with our own system.
Attached is what I've got. Basically if the number is negative [and highlighted yellow], the AWAY team wins. Which is a little confusing, because when you read your "home" record, yellow is negative, while when you read your "away" record, yellow is positive. But it is what it is.
It shows Purdue (in a double-round robin) with a 17-9 record, with home losses to MSU and Michigan, and away losses to IU, MD, M, MSU, UNL, OSU, and UW.
The new tier systems shows Purdue with a 19-7 record, with home losses to MSU and M, and away losses to M, MSU, MD, UNL, UW, IU, Iowa.
Essentially the only places it disagrees are Iowa and Ohio State. And it's close. It has Iowa as a 0.65 point dog over Purdue at Iowa. It has Ohio State as a 0.54 point favorite over Purdue at home. Given that Iowa is the weakest Tier 3 team we show, and Ohio State is the strongest [well, only] Tier 4 team we show, that looks pretty close to what we'd think.
It does have MSU as a 25-1 record. It actually only has MSU as a 0.11 point dog on the road at Michigan. Which for two teams both in Tier 1, seems like a decent gap.
It has Michigan at 24-2, though, also losing on the road at Nebraska, and VERY close against Purdue, with Michigan only a 0.25 point favorite in West Laf.
Just thought I'd pass this along. I've got the spreadsheet set up so that all I need to do is upload 14 numbers from the Sagarin rankings and the rest will automatically update.
If anyone believes 3.10 points is not a sufficient HCA in the B1G, though, let me know. I can change that number as well, and it'll probably make some significant changes to the outcome.
