I wasn't satisfied just comparing pitches over 100mph to pitches between 90-92, so I broke things down into something of a continuum. The results:
| Pitch Speed | Contact% | Foul/swing% |
| 90-92 | 84.9 | 40.1 |
| 92-94 | 83.5 | 42.3 |
| 94-96 | 81.8 | 43.6 |
| 96-98 | 81.5 | 44.5 |
| 98-100 | 74.2 | 43.5 |
| 100+ | 76.0 | 46.4 |
Once you get into the upper 90s and triple-digit range, you start dealing with sample size concerns, along with concerns over the matter that there are only a few guys who can throw that hard, potentially biasing the numbers. But what's abundantly clear - and what comes as little surprise - is that the harder a guy throws, the harder it is to square him up.
It's interesting to note the contact drop between 96-98 and 98-100, since we're still dealing with samples in the thousands.
0 recs | 8 comments
Would love to see this data combined someday with movement.
In other words, is 94 mph with six inches of vertical movement (Felix?) equal to or greater than the lack of contact at 102 mph straight ahead?
diderot - September 1, 2010
I suggested it in the other thread and I think there may be something to it
the 96-98 group still includes a fair number of starters while the 98-100 group includes almost none and the 100+ group is going to be just relievers and Strasburg. Relievers may have an unfair advantage in StrSw% on their fastballs on account of platoon matchups, entering late in the game (change of pace, etc) and competition.
I think limiting your sample to relief pitchers would do away with this concern as well as make the samples a little more equal. But I don’t think you should necessarily go and do that because you’ve already posted on this twice and I guess the result will still be the same only less pronounced (ie the drop from ~97mph to ~99mph would be more in line with the rest).
Bearskin Rugburn - September 1, 2010
I can't filter by starter/reliever anyway
That’s beyond my means.
Jeff Sullivan - September 1, 2010
Its interesting
fouls/contact (just (fouls/swing%) / (contact/swing%) which he has above) is pretty much completely linear with velocity.
Didn’t want to take the time to post the chart but:
fouls/contact = 0.0135*velocity – 0.7509
R^2 = 0.988
Edgar for Pres - September 1, 2010
Is a foul ball worth more than a swinging strike?
You can’t strike somebody out on a foul ball but foul balls can get caught for outs. Anybody know?
Edgar for Pres - September 1, 2010
Worth in what sense?
Swinging strikes are not worth anything per se, but correlate well with K rates. Strikeouts are worth something in terms of linear weights because they are one of several possible outcomes for an at bat but how they are achieved is not relevant to value.
In terms of value as the last event in an atbat a foul out or foul tip must have an identical value to a strikeout (out recorded, runners cannot possibly advance) though they are far rarer. I’m not sure it makes sense to assign value to pitch outcomes unless it is the final pitch of an atbat.
Bearskin Rugburn - September 1, 2010
I was thinking about how much a strike is worth from a run expectancy attitude
Going from a 0-0 count to 0-1 count on a hitter will decrease the chance a run scores on average.
Edgar for Pres - September 1, 2010
Not that I'd expect you to follow up on requests......
But at one point, Bill James looked at matchups between groundball/flyball pitchers and groundball/flyball hitters, and concluded that same type matchups (flyball vs flyball, groundball vs groundball) favored the pitcher, while opposite matchups favored the hitter. Somewhat like same-handed matchups. I’ve been curious what the newer information and datasets would have to say about this.
“This post is almost not off topic.”
nathaniel dawson - September 1, 2010
You must Login with your SB Nation account and be a member of Lookout Landing to post a comment.