Search Results for 'sagarin'

[shakeitsugaree]

You’re right – if Colorado runs the table they will undoubtedly be South champs – and rightfully so! Colorado’s two/three predicted losses, by system, are UW and Utah (Sagarin) or UW, Utah and USC (S&P+). If Colorado beats UW, Utah and USC they deserve to be South champs. Just like if Utah runs the table from here on out there will be quality wins (Stanford, USC and Colorado).

Thank you for the kind words about my posts! The reason I find these numbers interesting is because they are an attempt at removing bias from analysis. I especially like the S&P+ system because Connelly is so transparent in how he calculates it. He also has a fantastic SB Nation site, which really gets into the nitty gritty of statistical analysis.

The thing that has always bothered me is when people dismiss teams based on history or the ‘eye test.’ This is because I am a Utah fan and experienced one of the most epic upsets in college football history – which, btw, was predicted by an algorithm. Here’s an excerpt from an interesting read (find the rest of the article here)

The computers reclaimed center stage in 2008, when Massey’s ratings had undefeated Mountain West champ Utah ranked ahead of Alabama, which had been No. 1 in the human polls. At the time, it seemed ridiculous—even Massey agrees that it was borderline absurd. But, he says, once margin of victory went by the wayside, sometimes his poll results looked wonky. Alabama ended up facing Utah in the Sugar Bowl to end that season after losing to Florida in the SEC title game, and the Utes won handily, 31–17.

The problem with statistical analysis in football is sample size. This is why Utah’s predicted finish has been so volatile – there haven’t been enough events to really get a handle on this team’s capabilities. But that 7-2 record keeps coming up, that might be what we stabilize at. That would be enough to win the south, assuming Colorado loses to UW.

[Matthew Thomas Castleton]

Stanford would be 9-3, not 8-4 if they go 7-2 in conference. They went 2-1 in non- conference play (wins vs. San Diego State and UC Davis, loss @ Notre Dame), which would make their record 9-3 (7-2).

[UtMtBiker]

But if CU runs the table….. we’ve shown we are capable of s**tting the bed against CU. Not sure why,  but I enjoy your posts about these polls. I think they are total crap but it’s so fun to follow.  Go Utes! 

[shakeitsugaree]

That win over Stanford really increased our value to the prognosticators. Here are the updated in conference win predictions:

Sagarin
Utah 7-2 Pac12 South Champ (tie breaker is head to head win over Colorado)
USC 6-3
Colorado 7-2

S&P+
Utah 6-3
USC 7-2 Pac12 South Champ
Colorado 6-3

This is much improved from last week’s predicted 5-4 finish by both systems. Bottom line, we win the south if we win out.

[rbmw263]

stat models/vegas love us, and we really havent put it all together yet. The ceiling of this team remains very high.

[shakeitsugaree]

It was split; Sagarin predicted a loss, S&P+ a win. Both systems predicted wins over Colorado, before last week’s loss. Now the predictions are:

@ Furd L
v Zona W
v USC W (this is a bit misleading – S&P+ actually predicts a tie, so I gave us the W)
@ UCLA W
@ ASU W
v Ore W
@ Colo L

[Tony (admin)]

Did they predict us to lose at WSU?

[shakeitsugaree]

Well … that loss was, uhm, not good.

Both Sagarin and S&P+ predict we go 5-4 in conference this year. The real bummer is that we are predicted to beat SC, but lose to Colorado.

S&P+ predicts the following finishes: USC 7-2 *south champion by head-to-head win over Colorado*, Colorado 7-2, Utah 5-4

Sagarin: USC 7-2 *south champion*, Colorado 6-3, Utah 5-4

Wish I had better news.

[shakeitsugaree]

Short version: S&P+ says we go 7-2 and win the south; Sagarin says we go 6-3 and SC finishes 7-2.

One thing I’ve been thinking about this week is how much I like Connelly’s S&P+ metric. There’s been a lot of talk on this site about whether statistics are meaningful or useful in any way. I would recommend reading this explanation of S&P+ (I’ve linked to this before) before you decide to disregard statistics. You can find the S&P+ rankings here. Connelly claims that S&P+ tends to hit between 51 and 54 percent against the Vegas spread. Not bad. BTW, S&P+ says we win this Saturday; this is the difference in whether we take the south, because SC is also predicted to finish 7-2, but we would hold the head to head tiebreaker.

For a brief version of the long explanation, S&P+ evaluates a team’s performance in what Connelly defines as the Five Factors of winning: efficiency, explosiveness, field position, finishing drives, and turnovers. These factors are weighted to reflect their relative importance in predicting wins; Connelly states in his explanation that efficiency is the single most important stat in creating wins. What is efficiency? It is based on ‘success rate’ which Connelly defines thusly:

First downs: gaining at least 50 percent of necessary yardage (usually 5 yards) is successful.
Second downs: gaining at least 70 percent of necessary yardage is successful.
Third or fourth downs: gaining at least 100 percent of necessary yardage is successful.

There’s more to S&P+, but in looking at the breakdown in how each team is ranked, this system makes intuitive sense. For example, Utah as the #1 defense as determined by S&P+; Alabama has the #1 offense.

What do you guys think? Are there better analytics out there that anyone knows about?

[UtMtBiker]

We need help from WSU in Pullman in 2 weeks. Is there anyone who would bet on Utah to win that game?