Record keeping in college football is not the biggest issue facing the world. It isn’t now, it never has been, it never will be. It cannot create an existential crisis and it pales in comparison to anything that the world can truly call a problem. But record keeping in college football matters, and it falls short of what it should be. I wouldn’t go so far as to call it broken, but it needs to improve.
Some people who vote for major awards base their vote, in part, on statistics. The playoff selection committee were supposed to base their selections off of having watched the games, but they certainly justify their selections by using statistics, and they have given every appearance of having used statistics at least in part to make their decisions. Major awards are at stake, playoff appearances and thus people’s bonuses are at stake. Record keeping matters and those who do it should make every effort to get it right.
My original analysis (see link in previous post) led me to believe that Ohio State was missing on a lot of 3rd down opportunities, especially on 3rd and medium. This was based on this site (
HERE) which shows 3rd down passing statistics including the number of first downs gained.
In going through and watching every play, I discovered that the site that was the source of my data was making a critical error in their record keeping. For as long as there has been football, teams have been credited with first downs on plays on which they scored a touchdown. It is the long-established standard, but more than that, it makes sense. Third down percentage should indicate the % of the time that you were successful on third down; counting touchdowns as first downs ensures that. When I discovered discrepancies in the data I went through play-by-play records to ensure that it is still done this way and found that it is.
However, in compiling data for this analysis I discovered that the site in question only counted a touchdown as a first down if it was NOT a goal-to-go situation. For example, when Fields hit Wilson in the left corner of the end-zone to go up 4 scores against Cincinnati the Buckeyes were credited with a first down because it was 3rd and 8 from the 9 yard line (it was not a goal-to-go situation). But when Fields later found Hill in the end zone on 3rd and goal from the 5, the Buckeyes were not credited with a first down. The result for the Buckeyes is obvious: numerous 3rd and mediums were 3rd and goals, and every time the Buckeyes dropped back to pass on one of those, the site-in-question’s record-keeping practices recorded the effort as unsuccessful (first-down-wise), whether the Buckeyes scored or not.
Nor is the flawed record keeping restricted to the site in question. Ohio State’s own records were flawed:
- First downs gained in the first quarter vs Cincinnati were miscounted (may have been miscounted elsewhere, that’s the one I noticed)
- Play-by-play records recorded defensive players that never played a snap of offense as having been the target for passes on at least 5 occasions
- Play-by-play records recorded the pattern run by the receiver 3 times all season (the part of the field that was thrown to was recorded more often). All 3 times the pattern was called a post. All 3 times that was completely wrong.
- The first play of the fourth quarter vs Northwestern was recorded as a sack. Go watch it for yourself. Tackle for loss? Sure. Sack? No way.
The primary issue here though is Justin Fields’ efficiency on 3rd down. The flawed record keeping mentioned previously isn’t the reason he looked so good on 3rd down. In fact, the flawed methodology affected Justin more than any other quarterback in the country last year and he was actually even better than the previous analysis made him look. More on that in the next post.
