The role of stats, when used properly, is to provide an unbiased look at a player. Over time, we all become subject to our own prejudices, and thus we can reach conclusions that do not conform with reality. This is how the Lombos become Lombos. We decide something then form our opinions around that decision. We reinforce when we really should be questioning, constantly poking at our conclusion to make sure it's solid from all sides. The numbers can help you do that but only if you use them the right way.
One of the common refrains heard over the past couple years and especially this season is that Strasburg is a "hothouse flower". We've seen him break down after errors. We've heard him complain about mound conditions, about weather. He wilts. He collapses under pressure.
Ok that's the conclusion. But we got there the same way we've gotten to many places before, our own raw observation backed up by feelings and conjecture. What do the numbers say, or more precicsely what numbers would you like to look at to prove that there is an issue? Decide that first - see if the argument holds up.
If he wilts under pressure surely he'd collapse with runners in scoring position or in low-scoring games or in tight pressure situations late in games? Do we see that? Do we see a significantly worse performance post errors or do we not? Is this consistent over years or just this year? If we do see an issue, are we sure it's because of what we think or could it perhaps be about something else?
The latter question - one of confounding - is important. One of the current questions about Strasburg is why is he pitching worse away than at home. The numbers bear this out - looking at raw stats (2.41 ERA home, 5.25 away - corresponding fancy stats agree) or looking at patterns. (sort by game score and see the home and away games segregate). But is there something else afoot? Well yes, if you note Strasburg has pitched terribly with Sandy Leon as his catcher. Sandy Leon was catching for his two worst performances, both away games. This doesn't resolve the home/away question but it moderates it. Make Sandy Leon the issue and if you pull him out of the equation and the home away splits are closer; 2.50 to 4.09. The closer you get to even the more it becomes about random variation. This would make sense for a pitcher who's previous home/road splits show a very normal, very slight preference to home instead of a major issue.
This doesn't mean SOLVED! but it means what we were taking as a given requires a closer look. Frankly I do think the gap in the ERA and the segregation when you look at performance indicates something.
Ok so what about the OrKid? Well he's dead-on with RISP (.712 opponent OPS) compared to overall (.711), and great when there are 2-outs with RISP (.606). He's bad when the Nats don't score for him (.802), and hideous when it's "late and close" (.927), but scarily even worse when it's high leverage (.945). So maybe there's something here...
Yet last year he was worse with RISP and great "late and close". And that issue this year when the Nats don't score for him? Turns out this year it's this way for the entire team, making it seem like we're looking at the issue backwards. It's likely not that the Nats pitchers pitch worse when the Nats don't score, but the Nats stop scoring when staked to a big deficit. High leverage was also an issue for Strasburg in 2013, but not in 2012. It's tough to say there's a pattern of issues (or successes) here.
What about the error thing? His ERA in games where men reached base is 2.32. It certainly hasn't effected him in later innings. What about in the moment? Guys batting in the same inning post error this year have hit .278 with one XBH. Nothing at all strange here. But yet it persists.
The work I've just done is admittedly cursory but trying to be objective we've found little to suggest some sort of internal weakness. I suppose you could work it as - ok THIS YEAR he's mentally fragile and it may not show in all standard high-pressure sitiations but when they pressure is REALLY on he breaks - but that strikes me as fitting the data to your conclusion, rather than seeing what conclusions can be reached with the data.
Strasburg may in fact have an issue - but you don't know, can't prove, and
can't even really see when looking at these past few years as a whole outside of "well I can read his body language". There is a disconnect between the talent and the result but trying to pin down a mental "fragility" issue does no one any good. If it is a fastball issue (speed? location? more likely a combination of both?) then that's what the focus should be on because the speed isn't coming back, not fixing his head. Especially not when the diagnosis is based on little more than what you've seen while lounging on your armchair.
*Some more interesting Strasburg stat diversions for those inclined - following up the stuff I was looking at earlier in the season with his 2-strike issues (damn that target data isn't readily available). - what has changed about the Nats' pitching approach? Renewed focus on keeping runners honest. What's new in Strasburg's stats this year? An issue with men on first. How do we prove/disprove this? some sort of speed of runner adjustment combined with if the SB would matter score wise? Would the subsequent PAs be enough to really make a judgment?