BFS,
There’s a lot of things that can be said for the MaxPreps stats, but accuracy to the level of MLB is definitely not one of them.
The 1st thing to understand about MP and other services like it is, the data is only as good as the SK and/or anyone else who has the password to enter and/or edit the data. I don’t know for sure, but after having communicated with literally hundreds of HCs about it, I think its safe to say most MP stats are entered manually by a coach, but that’s slowly changing as apps like IS and GC become more widely used. Here’s what happens when stats are manually entered.
Someone has to either have the scorebook in front of them and total every entry as they’re putting in the data, or they have to have totaled everything for every player ahead of time. But either way it presents a problem because there’s human beings involved, and we all make mistakes doing both totaling the data and doing the actual data entry.
The 1st thing MP wants is the final score. Its only necessary to enter the score, but the score by inning and hits and errors. And this brings up the first thing that “might” be an issue. Our league as well as many others in the area have Arbiter enter the schedules and scores, so whoever checks the data has to make sure the schedule’s correct and the score is correct. When I enter stats, I usually don’t do it until the next day after I’ve double checked everything, so in almost every case the score is already posted, but only the score is entered, so I have to put in the runs by inning and the hits and errors. And that brings up another little “problem”.
Either team can change the game score, hits and errors. So if Arbiter enters a score of 8-3, and I edit it to show the runs by inning as 5-3-0-0-0-0-0 with 9 hits and 2 errors, and 1-1-0-0-0-1-0 with 5 hits and 4 errors, it isn’t necessarily over. Anyone with a password for our school or the opponent’s school can change either the score or the hits and errors! That really wouldn’t be a big deal, but those are the things the standings are automatically made up from, and that the folks who do the rankings use to do that. I’ve caught a few people making changes to scores to enhance their ranking, but its pretty rare. Usually there’s only difference in hits and errors, not scores.
Once the score and R-H are entered, you can start entering stats for the players. The user is presented with the roster in uniform # order, and the 1st thing that can be entered is the 17 batting and running stats. Each one has to be entered individually, so if you had 12 different players hit or run, that’s 204 possible entries. Here’s the bad part. There’s no data verification done. For example, its possible to enter 6 hits for a player who only had 3 ABs. In my program I check a lot of things like that so the bad data can’t be put in. The reason that’s important is, many people are like myself and use the number pad. Well, the 6 is less than a 1/16th of an inch from the 3, so “fat-finger” mistakes aren’t at all unusual.
Once the offensive data has been put in, you hit a button and the data is entered into the database, and the user’s taken automatically to the defensive data input screen. Here again the roster sorted by uni # is presented, and 7 different pieces of data can be input. After the data has been entered into the database, the pitcher’s data screen is presented, made up of 26 different data points. Once its all been entered, another button is press and the pitcher’s data hits the database. Here’s where I’m a lot different than “most”.
I’ve already made sure my data and totals are correct, so I go back into the hitting, defense, and pitching screens to check the totals MP has with the totals I have. That can’t be done during the data entry because MP doesn’t total anything until its put into the database. So if someone putting in the data doesn’t check very closely as its being entered, and they don’t go back and check totals, bad data is in there.
I have a standard spiel I give to coaches, players, and parents, that when they see something they think is wrong, to let me know ASAP. Either they’re wrong and I’ll give them an explanation why that is, or I’m wrong and I’ll go in and change it.
So, you can see how its very possible data might be accidently bad, or in some case purposely bad. It should also explain why some schools don’t put in “ALL” the possible data. Its not like it takes hours, but it’s not a 2 minute job either. MP doesn’t require any field be filled, nor does it require data for every game. What happens is, schools like ours who have someone willing to do it and the time to do it, usually put in everything they can, while others where the coach has to do it, it takes a much lower priority.
The nice thing about these scoring apps, is they export the data, so after the score is entered, instead of entering the data manually, all that has to be done is to hit the import key and its done automatically. But here again is where more problems can happen. I love importing the data, but even though I know its gonna work, I check all the totals against what I have.
Now comes the trouble with the apps. They’re great and I encourage people to use them, but it doesn’t matter how easy they are to use, if the person using them doesn’t know how to keep score well, its even easier to get bad data into the system.
So, the answer to your original question, are the MaxPreps stats for the most part worthless, is it depends on the school. For the individual school, they’re pretty good because they’re usually consistent, so you can pick out who has the bet BA or ERA on the team and be confident it shows good comparisons. But as much as I like MP and know the folks in our league generally do a pretty fair job of scoring and putting in data, personally I don’t do a lot of comparing players using the MP stats other than look at HRs for hitters or Ks for pitchers. And that goes for sure about trying to compare players from even larger pools like a division or state.
But the good news is, even with all the problems in MP, the data is immensely better than even just 5 years ago.