Bill James on Davey Lopes
More from 1979
Bill James on Davey Lopes
It’s been a few weeks since we last looked at a Bill James Baseball Abstract quote. Here’s another one from this summary of the 1979 Abstract:
This is actually an interesting discussion for those of us who play baseball sims. In games like APBA, the player’s speed rating is not really an indication of the player’s raw speed, but, rather, his ability to score when he is on base. If Lopes indeed scored fewer runs than he ought to have scored in 1978, we’d probably see him receive no speed rating, or even a slow speed rating.
He actually received a fast rating:
The next thing to check here is whether Lopes actually scored fewer runs than we’d expect. Actually, we don’t really need to use James’ contrived and arbitrary formula to figure this out. There’s a much simpler way to get a ballpark feel for it.
I grabbed National League hitting stats from Baseball Reference and created a “time on base” estimate the easiest way I knew how. I simply multiplied on base percentage by plate appearances. Now, the number doesn’t come out nice and clean (probably due to rounding), but it’s close enough for a real quick study.
I then divided the number of runs scored by the number of times on base, and expressed it as a percentage — again, rounded up. It doesn’t work for a few players who somehow scored more runs than their combined number of walks, hits, intentional walks, and times reacing due to error — I’m guessing pinch running had something to do with it. It’s not a huge deal, though, for something I’m just slapping together.
Anyway, here’s what the leaders in times on base looked like in the 1978 National League, as well as the percentage of the time they scored:
On average, players in the 1978 National League seem to have scored a run 32.50% of the time they were on base.
In other words, I think Bill James was trying to make a mountain out of a molehill here. I’m not seeing any reason to think that Lopes scored fewer runs than expected in 1978.
It’s not like Davey was a singles hitter, either. Loopes hit 17 home runs in 1978 and 25 doubles, meaning that he had ample chances to score. He also stole 45 bases and was only caught 4 times, which, honestly, is quite impressive.
James’ other problem, of course, is that Bill North didn’t hit second in the lineup all that often in 1978. It was Bill Russell most of the time:
James probably could have figured this out if he paid attention to the Dodgers before October, or if he actually went through the boxscores to tally things up as he allegedly did in those days.
North, by the way, only had 10 RBIs for the Dodgers in 1978, which was largely due to his awful .266 slugging percentage. He started to hit second in the lineup with some regularity on July 19, but Russell returned to the number 2 spot on August 3.
In short:
Lopes did not score fewer runs than expected in 1978
North’s bad hitting had little to do with how many runs Lopes scored
Bill Russell hit second in the lineup far more frequently than North did
Sorry, Bill.
Not a fan of APBA, but their 'fast rating' seems more than appropriate, based on extra base hits (4 triples) &, of course, his stolen base/cs numbers.
How else to gauge/measure players from the past per replay sims? Stolen base numbers can be attributed to the strategies of the times, but still...stealing a base does require some speed.
A fan of Bill James; but acknowledge his desire to dig deep (too deep?) to create a formula for everything. :)
Of course, baseball is one of those rare sports that can lend itself to doing so; making it one the 'easiest' sports to sim with a reasonable amount of accuracy. imho
Considering that I'm pretty sure we never saw that formula once he got a publisher a few years later, I think he realized its inadequacy soon after that. But his lineup error is disconcerting; he did need to look at every box score to analyze the pitching if nothing else. And Russell did bat second at least as often as North did in the postseason, so I doubt that he was extrapolating from that.