First, here’s a look at the averages for each metric. A couple of things stand out – the plays are almost equal, with winners running a paltry two more plays per game. It’s what they do with those plays (Yards/Play) that matters. Secondly, many people say total yards don’t matter. They do. More on this in a minute.

Here’s a look at home vs. away and favorites vs. underdogs. 56% seems a bit low given that most Power 5 teams have 1 or 2 “gimmee” games on their schedule. When you look at conference games this number is traditionally much closer to 50%. More important than playing at home is being the favorite. The old adage “the best team usually wins” is in large part true, assuming the best team is favored.

These numbers show the % of teams that have the better number for each category. For instance, the winning team has more plays in 54.8% of the games, more total yards in 77.7% of the games (told you yards were important) and higher yards per play in 79.6% of the games. Yards per play and total yards are the two most important stats in my book. Secondly, we often hear about “winning” the turnover battle. Since 22.9% of the time the turnover battle is even it’s more important to not lose the turnover battle (win or be even) – 82.3% of winning teams are at least even on turnovers.

]]>Only three teams had success rates higher than Clemson for all plays – two of them you can easily guess, Georgia and Georgia Tech. The third – North Carolina – is a little surprising considering the day Deshaun Watson had.

The criteria for a successful play are as follows: 50% of yards needed for a 1st down or a touchdown on 1st down plays, or 70% of yards needed for a first down or touchdown on 2nd down, or a first down or touchdown on 3rd and 4th down plays.

In an upcoming post I will merge these and the Clemson offensive success rates together on the same graph to give a game by game view of each, but for now here are the game by game rates for all plays, rushing (excludes kneel down plays and does not include sacks) and passing (passes plus sacks) for Clemson opponents in 2014.Obviously, the lower the number the better.

**All Plays**

**Rushing Plays**

**Passing Plays**

Offensive rebound percentage is an estimate of the percentage of available offensive rebounds a player got while he was on the floor.

]]>The eyeball test shows the obvious, but these numbers help to quantify that difference. Perhaps the most telling is yards per play while in the game. Clemson averaged 2.3 (2.20 when taken to the second decimal) more yards per play with Watson at quarterback than with Stoudt.

Related to that metric is the percentage of explosive plays. A difference of 4.7% may not seem like a lot, but over the course of a season (911 snaps) that would come out to 43 more explosive plays with Watson – about 3.5 per game.

In a coming post I will show the difference in Clemson’s win probability vs. Oklahoma in 3 different scenarios: (1) With the quarterback duties shared as they have been this season; (2) with Watson alone and (3) with Stoudt alone, and in the process quantifying the expected difference in the teams performance with each scenario.

* Please note that the Yard Per Play by Game (last two graphs) contain different scales.

]]>It looks like the Tigers lost one they should have won (Florida State) and won one the could have easily lost (Louisville). Most of the others are pretty cut and dried (and obvious) except for Wake Forest. How’d the model come up with 100% win probability for a game tied at 20 with 11:08 to go?

Actually, it’s pretty simple. We’re looking at the game retroactively, without the benefit of the ebbs and flows of the game. In the 2,801 games I’ve tracked no team has lost with the metrics advantages Clemson had at the end of the game. But even at the point the game was tied with 11:08 to go the metrics gave Clemson a 94.1% probability of winning because the Tigers were dominating the total yards and yards per play metrics.

Sure enough, the first play from scrimmage after the game tying Wake field goal saw Artavis Scott go 68 yards for a touchdown on the jet sweep/pass. Wake Forest was held to 0 yards on the next drive and then Clemson drove 72 yards on 9 plays to go up 14 and end the suspense.

From the point it was tied at 20 until the end of the game Clemson out gained Wake 157 to -17 and is why, in the end, the probability reached 100%.

]]>

The criteria for a successful play are as follows: 50% of yards needed for a 1st down or a touchdown on 1st down plays, or 70% of yards needed for a first down or touchdown on 2nd down, or a first down or touchdown on 3rd and 4th down plays.

Here are the game by game rates for rushing (excludes kneel down plays and does not include sacks) and passing (passes plus sacks) for Clemson in 2014.

**All Plays**

**Rushing**

**Passing**