So this week we are going to try something new. I am a proponent of the "football analytics" movement and would like to bring some of those advanced stats to you guys. I'll start off with an advanced version of the typical box score you would find on ESPN. The box score tells you the total yards gained by each team, how each team did running the ball, how they did passing the ball, tells you about penalties and turnovers, and the time of possession. All interesting information. But I think we can do better. Here is a version of the Advanced Box Score that Bill C has used in the past.
|Passing Downs||Quarter Performance|
|Num Plays||18||18||Yards Per Play (# of plays)||Success Rate||Yards Per Play||Success Rate|
|Success Rate||50%||28%||30%||Q1||7.21 (14)||43%||6.09 (22)||45%|
|Avg Yards||6.78||4.38||5.72||Q2||7.09 (23)||57%||2.44 (16)||13%|
|Standard Downs||Q3||4.94 (18)||56%||3.4 (15)||40%|
|GT||Duke||Natl Average||Q4||5.52 (21)||57%||3.33 (9)||44%|
|Num Plays||58||44||Down Performance|
|Avg Yards||5.98||3.97||5.65||Yards Per Play||Success Rate||Yards Per Play||Success Rate|
|Avg Yards||7.81||3.63||6.47||Drive Performance|
|SD/PD Sack Rate||0% / 0%||2.3% / 5.6%||1.8% / 4.2%||Num of Possesions||Avg Start Spot||Explosive Drive %||Methodical Drive %|
|Num Plays||60||32||% of Possible Yards Gained||GT: 61%||Duke: 28%|
|Success Rate||55%||38%||43%||Georgia Tech Hidden Yards||73.7|
|Avg Yards||5.73||4.53||4.95||Turnovers||GT: 1||Duke: 0|
I'll go through this box score item by item to explain everything.
- Success Rate: Success rate tells you wether or not the offense "did its job" on each play. A play is a success on 1st down if it gains more than 50% of the yardage necessary for a first down, on 2nd down an offense needs to gain 70% to be successful, and on third and fourth down an offense just needs to get a first down (or score a touchdown).
- Passing Downs: Passing Downs are defined as all 2nd and 7's or more, 3rd and 5's or more, and 4th and 5's or more plays an offense encounters. GT and Duke both had 18 such plays that would be considered "passing downs". On those plays GT averaged 6.78 yards per play, either running or passing, while Duke only averaged 4.38 yards per play, compared to the 2013 average of 5.72 yards per play on passing downs.
- Standard Downs: Standard Downs are all downs that aren't passing downs (all first downs are standard downs). GT's offense exceeded the national average against Duke and outperformed Duke's offense.
- Passing and Running Plays: These are the actual plays the offense passed the ball (or got sacked) and ran the ball (or scrambled after dropping back). In addition to average yards the Standard Down and Passing Down Sack rates are included (SD / PD Sack Rate). This is broken down for each offense so GT didn't get sacked on the day and had a 0% sack rate.
- Quarter Performance and Down Performance: How each offense did in each quarter by the same measures and also by down.
- Drive Performance: These are some of my favorites. Explosive drives are any drive that averages at least 10 yards per play. Georgia Tech had 11 drives (excluding the two at the end of each half where we let the clock run out, thanks for pointing this out dhbartlett!) and 2 of them were "explosive". Methodical drives are drives that last at least 10 plays. "% of Possible Yards" is the amount of yards an offense gained that were available to them. So a touchdown drive gains 100% of the yards available to an offense on that drive. Georgia Tech clearly outperformed Duke in this category. Hidden yards are derived from GT's average start spot advantage over 11 drives. So Georgia Tech started about about 6 yards closer to the endzone on average, so our hidden yards was 73.7.
So what are some advantages of this box score over the traditional one? Well, it tells you how each offense did in each category. Was their success gained from passing or running? Did a team have a field position advantage or did they drive the ball well? Here are some of my observations from this breakdown over a traditional box score.
- Our 3rd down offense was OFF DA CHAIN. The national average 3rd down success rate this year is right at 40%. Georgia Tech was at 65%. Nearly two thirds of our 3rd downs were successful. Normally I would say this isn't sustainable, however Georgia Tech's average yards to go on third downs was only 5 yards to go (national average of 6.93 yards to go). Our success on 1st and 2nd down should help us continue our success on third down.
- Our defense did a great job all around, but also did an outstanding job on passing downs. Granted, Duke had a backup QB in who looked lost under any pressure, but we still did a great job with everything.
- Duke helped themselves as much as they could with field position, I would have expected us to have many more hidden yards than we did, I'm just glad we were able to move the ball with such long drives ahead of us.
There are obviously some things that this box score misses (I will try and add a special teams category and a penalty breakdown in the future) but I think it does a great job explaining what happened in the game. And the best part? This isn't even as good as it can be. Average yards per play is a good measure of an offense, but not the best. Adjusted Net Yards Per Attempt takes into account touchdown plays and turnovers. For instance GT's average yards per pass was 6.8, but our AdjNYA takes into account the four touchdowns Vad threw (with a 20 yard "bonus") and the one pick (with a 45 yard penalty) to give the offense an AdjNYA of 10.1 yards per play. The national average AdjNYA is only 5.76 yards per pass, so our passing offense KILLED IT this week. And even beyond that, Bill C's Equivalent Points assigns a point value to each yard line to get a better sense of the overall worth of an offensive play. I will be doing these for each conference game and would like to improve it and introduce new things as the season goes on, but I wanted to get some feedback as well. So, what do you guys think? Too much information? Not enough? Anything else you would like to see? Let me know in the comments.