Numbers behind the turnaround

The Chicago White Sox acquired James Shields in early June with the intention of upgrading their pitching staff. They were just three games behind the first-place Indians at the time, with a record of 29–28. “We’re pleased to add a starter of James Shields’ caliber,” White Sox general manager Rick Hahn said.

Shields proceeded to have three miserable starts in a row for his new club, allowing 14 runs in a total of 7 innings the first two starts, and following that up with an 8-run, 1.2 IP effort against the Indians in his next start.

On top of that, Shields had allowed 10 runs over 2.2 innings pitched in his final game with the Padres. So, over a 4-game stretch between May 31 and June 18, Shields had pitched 11.1 Innings and allowed 32 runs.

Then, seemingly out of nowhere, Shields went back to being a dependable starting pitcher again. He has now strung together seven straight outings without allowing more than three runs in any of them. The totals since that tough four-game stretch: 47 Innings; 11 runs allowed.

So, how do you explain a turnaround like that? We may not have all of the answers, but we do have some interesting data to look at. Let’s get to it! We’ll call the 4-game stretch “Bad Shields” and the 7-game run he’s on since then “Good Shields.”


  1. Behind or Even in the Count: During Shields’ four-game meltdown, hitters were 20-for-37 (.541 BAVG) with 6 home runs when he fell behind or had an even count against them. Good Shields is allowing very little damage on those counts, and throwing more strikes to boot.

2. Left-Handed Batters: Shields found it difficult to get lefties out during his tough stretch. Since then, he’s been lights out against them. Left-handed batters are just 13-for-82 against “Good Shields.” Most impressively, they are just 1-for-29 against his fastball in his last seven games. They were 6-for-10 against it when Bad Shields was on the mound.

3. Early Strikes; Early Outs: One thing Shields has not done during his seven-game turnaround is strike hitters out. Good Shields has a strikeout rate of just 12.6%, far below the league average 20.9%. But he has been able to let hitters get themselves out early in the count. How’s this for efficiency?: Opposing hitters are 2-for-25 (.080 batting average) on the first pitch of the AB during Shields’ last seven games. The league average on first pitch is .346.

The table below shows that 49.5% of the plate appearances against Good Shields have been over with by the third pitch, compared to just 42.5% for Bad Shields. More importantly, 56% of the outs that Good Shields has recorded have been in three pitches or less. That’s the fifth-highest rate for any starting pitcher who has made at least 5 starts during that timeframe (League Avg: 46%). And the outs have typically been weak or medium contact. Only 8 have been well-hit.

4. Edge fastballs called strikes: We classify pitches that are within 1 baseball width of the strike zone (in or out of the zone) as “pitch near the edge.” Shields has not thrown more “near the edge” fastballs during his seven-game hot streak, but he has been getting more calls. For Bad Shields, only 53% of the taken fastballs that were near the edge were called strikes. Since then, it’s 70% (37-of-53). That’s well above the league average 60%.

Just one example of an “edge” fastball taken for a strike

5. The return of the changeup: Shields’ best off-speed pitch is his changeup — a pitch that turned sour on him during the Bad Shields games. It’s now sweet again.

So, there you have it. Pitching to contact, getting the benefit of some calls, and finding his changeup again. That pretty much sums up the recent turnaround of James Shields.

Please take a moment to hit the heart button if you like the content. We’d really appreciate it!

You can follow Inside Edge on Twitter @InsideEdgeScout