Baserunning Ratings in Baseball Sims, Part 2. Strat-O-Matic

[@EricBerinson]

Hi Steve, Great part 2 follow-up. Your material always makes me think! Baseball reference appears to provide 2 stats that together seem to do a great job exposing pure baserunning (NOT base stealing) performance: One- XBT: Advancing an extra base on a single or double 1a- You pointed out “XBT,” % times runner advanced an extra base from 1st or 2nd 1b- An XBT% is given at the league as well as player level- LgXBT% 1c- XBT+ like OPS+ would be player’s XBT vs. league avg: 100 x (XBT% / Lg XBT%) Two- BT: Advancing a base in non-hit scenarios 2a- BT are Bases Taken on fly balls, passed balls, wp, balks & def indifference. 2b- BT opps are H + BB + HBP + ROE (Reach on Err) + XI (on per catcher int.) - HR 2c- Both BT & BT opps are accessible for league and player at Baseball Reference 2d- So, BT% = BT / BT opps 2e- BT+ like OPS+ is players BT% vs league average: 100 x (BT% / Lg BT%) For simplicity, the average of XBT+ and BT+ provides BR+ (baserunning rating like the OPS+ scale) that’s pretty complete for baserunning. One could weight these, I am suggesting to start they (XBT+ and BT+) just carry equal weight? What’s missing for pure baserunning? Not much! Successful scoring on a Sac Fly is included in #2, along with so many other non-base hit baserunning scenarios. One does comes to mind: Picked off when NOT attempting to steal (somewhat subjective). Baseball Ref does have some pickoff stats, but not sure they could be used here (too related to SB attempts). Still, do you concur BR+ is a potential solid pure baserunning stat for sim ratings? The Pudner formula is excellent, but uses triples, SB’s and runs scored. Baserunning clearly correlates to these 3 speed related stats, but it’s also a bit independent of them too. I will be collecting and comparing Strat-O-Matic Running ratings alongside BR+, Bill James speed score, and a modified Bill James speed score (including BR+ as a 7th component) to gauge possible correlation. I’d love to hear comments if my suggestion is flawed, or can be improved? Thanks!

[@steve_etzel]

@@EricBerinson Hi Eric. Thanks for your excellent response! I think your suggested approach will produce better results than the oversimplified Pudner approach. The underlying stats appear to measure exactly what a baserunning rating is used for. I would be interested in knowing how your stats correlate with baserunning ratings from Strat and other sims. When I’m making homebrew cards for sims, I usually co-opt the Strat ratings and use the Pudner formula as a fallback. I played around with the James speed rating but that didn’t provide me with much more info than what I already had. I think I emphasized the Pudner formula due to laziness. I didn’t want to take the extra effort to dig out those supplemental running stats. Thanks again.

[@steve_etzel]

@@EricBerinson One other thought. I wouldn’t be too concerned about pickoffs. They are pretty rare. If I were to include them somewhere it would probably be a component of a stealing rating, as a pickoff is essentially a caught stealing. For pitchers, I include pickoffs when calculating a Hold rating. Noting however, that hold rating calculations need to be customized to the specific sim rules for which the rating will be used. (Does the rating affect SB attempts or success rate, or both?)

[@steve_etzel]

I haven't studied this a lot, but here is another article on the subject: charliepavitt.home.blog/wp-content/uploads/2018/10/baserunning.odt

[@larryrice5231]

@EricBerinson This is cool. I see the Bases Taken stats in BB Ref but I don't see the BT Opportunities. Would this have to be separately calculated?

[@EricBerinson]

@@larryrice5231 Hi Larry, sorry I was using a bit of shorthand in my reply so it may not have been clear. BT opportunities are described in my "step 2b:" H + BB + HBP + ROE (Reach on Err) + XI (on per catcher int.) - HR. This info is available on 2 different Baseball Ref spots for a hitter... You can get ROE and XI for a player and for the league for any season using the web site Steve had alluded to in his presentation (or just search "Baseball Ref Baserunning and Misc."). The Season league summary is in the specific columns (ROE, XI) for both those is at the top of the page under the team totals, and player summaries for that season follow. You can get H, BB, HR and HBP for a player and for the league for any season by searching for "Baseball Ref Standard Batting..." Again, the Season league summary for the 4 of these stats is in the specific (H, BB, HR and HBP) columns at the top of the page under the team totals, and player summaries for that season follow.

[@larryrice5231]

Thanks for laying this out Steve. I really need to check out the last references in your presentation. I’d really like to experiment perhaps with something different than I did previously on BU card making. I guess now would be the time…

[@steve_etzel]

@@larryrice5231 Right now the BU spreadsheet has a version of this formula as the default. Some of the Fangraphs stats could be useful. I think what you and I have been doing by using other sources has worked ok too.

[@johnnysmoke612]

As always enjoyed your presentation. I speculate SOM special ratings are never fully objective but objective with some subjective adjustments for overall and team stats accuracy.

[@steve_etzel]

@@johnnysmoke612 From everything that I have read, I believe you are correct. Numbers may play a part (probably more now than in the past) but ratings are debated by a group before being finalized.

[@johnnysmoke612]

@steve_etzel I've always thought SOM does a real good job with their subjective ratings. I highly admire the game engine but fielder errors by bases empty versus runners on base, on pick offs, and on bunts need much correction and work. I've compared the engine versus real life data and found correlation errors for the lack of the proper error off the proper result situation. Today's advanced data availability SOM should tweak their engine for correcting these short comings. It doesn't have to be done by assigning specific e-ratings for each player for each type situation like t-ratings for catchers on stolen bases. But could be done by letting the e-rating in the overall work for the given situation with the error. For example, 75% of all outfielder errors are off bobbles or throws after fielding a hit. A outfielder with a e-rating of 10 would give up 7 or 8 hits off a different outfielder x-chart for when an outfielder e-rating is rolled. In this case, it works very well just using the overall e-ratings since 70-75% of outfielder errors are the larger sample. Links for chart and instructions. www.mediafire.com/file/gfqvw6g43y6gced/Outfield_Error__X_Chart_E_Rating_Procedures.txt/file www.mediafire.com/file/uz6oze3bk7apl4t/Outfield_Error_X_Chart_PDF.pdf/file