Poll 2017: Which Group of Pitchers Performs Better? — A Review

October 3, 2017

Once again, heading into the all-star break, I presented two lists of 10 pitchers each — both lists composed of pitchers on each side of SIERA-ERA differential. I then asked you to vote on which group, the SIERA overperformers or underperformers, would power a lower second half ERA, and which ERA range each group will land in as a group average. Amazingly, the results suggested these two groups of pitchers were identical! Both garnered between 44% and 45% of the vote to post the lower ERA, while nearly 11.5% of you couldn’t choose a side, predicting the groups would finish within .05 earned runs of each other. Furthermore, 29% voted that 4.00-4.24 would be the ERA range that both groups would settle into, with 3.75-3.99 the second highest vote-getter at 27% and 28%. Pretty crazy how close these votes were! Let’s see what actually happened.

All stats are from the second half:

Group A – The SIERA Outperformers

Name	K%	BB%	LD%	GB%	FB%	IFFB%	BABIP	LOB%	HR/FB	ERA	SIERA	ERA-SIERA
Andrew Cashner	12.9%	7.8%	18.6%	47.4%	33.9%	15.1%	0.250	73.2%	9.7%	3.27	5.33	-2.06
Brandon McCarthy*
Chase Anderson	23.3%	6.8%	16.5%	41.0%	42.4%	8.5%	0.252	84.4%	10.2%	2.47	4.09	-1.62
Dallas Keuchel	18.6%	9.6%	16.6%	66.4%	17.1%	0.0%	0.286	73.8%	25.0%	4.24	4.09	0.15
Ervin Santana	20.1%	5.5%	17.8%	37.3%	44.9%	11.3%	0.279	74.6%	10.5%	3.66	4.40	-0.74
Gio Gonzalez	21.8%	8.4%	18.6%	47.9%	33.5%	6.2%	0.255	77.0%	8.6%	3.08	4.32	-1.24
Ivan Nova	21.2%	7.0%	24.4%	41.6%	34.0%	9.9%	0.360	65.3%	19.7%	5.83	4.29	1.54
Jason Vargas	17.1%	10.1%	19.9%	44.1%	36.0%	11.8%	0.309	69.7%	18.8%	6.38	5.17	1.21
Jose Urena	15.4%	9.5%	19.1%	46.7%	34.2%	6.8%	0.252	79.9%	15.9%	4.10	5.25	-1.15
Kyle Freeland	18.8%	9.8%	21.7%	51.3%	27.0%	7.3%	0.355	69.3%	12.2%	4.81	4.67	0.14
Group Average	18.5%	8.2%	19.2%	46.9%	33.9%	8.7%	0.286	74.0%	14.5%	4.16	4.66	-0.49
League Average (All Starters)	20.8%	8.1%	20.8%	43.7%	35.5%	9.5%	0.300	72.0%	14.0%	4.50	4.47	0.03

-All Group Average metrics (excluding ERA & SIERA) were calculated by weighting total batters faced, which isn’t perfectly accurate, as the batted ball data should be weighted based on balls in play, but that would have been too time-consuming; these averages are close enough
*McCarthy didn’t pitch in the second half

Group A – The SIERA Underperformers

Name	K%	BB%	LD%	GB%	FB%	IFFB%	BABIP	LOB%	HR/FB	ERA	SIERA	ERA-SIERA
Adam Wainwright	6.0%	10.3%	25.5%	45.7%	28.7%	3.7%	0.261	73.5%	14.8%	4.78	6.47	-1.69
Jeff Samardzija	21.6%	5.0%	18.7%	39.3%	42.0%	9.1%	0.278	68.1%	10.9%	4.22	4.11	0.11
Josh Tomlin	21.2%	1.6%	20.0%	40.0%	40.0%	10.7%	0.289	78.8%	10.7%	3.19	3.82	-0.63
Kevin Gausman	26.2%	7.7%	19.7%	44.5%	35.7%	8.2%	0.289	85.4%	17.6%	3.41	3.80	-0.39
Kyle Gibson	22.1%	6.9%	24.1%	50.9%	25.0%	9.1%	0.319	77.1%	14.5%	3.76	3.93	-0.17
Marco Estrada	18.1%	8.8%	17.8%	28.4%	53.8%	21.1%	0.262	71.7%	9.9%	4.76	5.29	-0.53
Masahiro Tanaka	29.6%	4.6%	18.9%	49.8%	31.3%	11.1%	0.289	72.1%	19.0%	3.77	3.03	0.74
Matt Moore	19.9%	7.8%	19.8%	38.3%	41.9%	14.7%	0.281	68.0%	12.6%	4.86	4.63	0.23
Trevor Bauer	26.7%	7.5%	22.5%	45.4%	32.2%	11.0%	0.338	87.1%	15.1%	3.01	3.69	-0.68
Ubaldo Jimenez	24.5%	5.8%	23.4%	41.0%	35.6%	4.5%	0.408	68.7%	20.9%	7.02	3.81	3.21
Group Average	21.6%	6.6%	21.1%	42.7%	36.2%	10.3%	0.301	75.9%	14.8%	4.19	4.13	0.07
League Average (All Starters)	20.8%	8.1%	20.8%	43.7%	35.5%	9.5%	0.300	72.0%	14.0%	4.50	4.47	0.03

*All Group Average metrics (excluding ERA & SIERA) were calculated by weighting total batters faced, which isn’t perfectly accurate, as the batted ball data should be weighted based on balls in play, but that would have been too time-consuming; these averages are close enough

Would you look at that, it was essentially a dead heat! Group A outperformed Group B by the slimmest of margins, posting an ERA just .03 earned runs better, making the 11.5% that voted for a near tie the victors.

In the first half, Group B posted stronger skills as per SIERA, while posting a significantly better strikeout rate. The second half was no different, as Group A’s skills were just as bad, while Group B improved theirs slightly, and also increased their strikeout rate even higher. Once again, Group A suppressed line drives better than Group B, which most certainly helps explains the meaningful 0.15 gap in BABIP. This is even more interesting considering Group A induced more grounders and fewer pop-ups, both of which should boost BABIP, not reduce it.

Both groups posted near identical HR/FB rates, while Group B actually stranded a higher percentage of runners (LOB%). Both of those metrics totally reversed course from the first half. During that period, A posted a ridiculous 80.5% LOB% to B’s lowly 68.1% mark, while A kept their fly balls in the park to the tune of an 11% HR/FB rate, while B didn’t enjoy such luck, posting a 16% mark. This is precisely why we call these the “luck metrics”. Over small samples, fortune and randomness play a large role. But more often than not, they even themselves out, which is what we saw in the second half.

It’s doubtful fantasy owners will be opening their wallets for Andrew Cashner next season, but Chase Anderson is a name whose breakout many are going to believe. Don’t be that person. He was one of the most fortunate pitchers in baseball this year and owns a SIERA above 4.00 for the season. Meanwhile, Kevin Gausman figured out how to be good again in the second half, but his season line still looks ugly. He’ll either be a trendy sleeper that pushes up his price or a bargain. And can you believe Ubaldo Jimenez posted a sub-4.00 SIERA in the second half?! It’s too bad seemingly every ball in play fell for a hit and fly ball sailed over the wall.

BAL	CHW	LAA
BOS	CLE	OAK
NYY	DET	SEA
TBR	KCR	TEX
TOR	MIN	HOU

ATL	CHC*	ARI
MIA	CIN	COL
WSN	MIL	LAD
NYM*	PIT	SDP*
PHI	STL	SFG