Meant to Be? The Rockies and the 3-3-3 Rotation

Since the Rockies have started playing baseball in Colorado, they’ve continually run into the same problem: pitching. We’re all familiar with the situation — the altitude and thin air create a hitter’s haven and a nightmare for pitchers, particularly starting pitchers. The Rockies have tried to remedy the situation in the past by bringing in top-tier starting pitchers, only to have them struggle. In 2012 and ’13 they tried a four-man rotation with a 75-pitch limit which led to a 64-98 record and a 5.22 team ERA. 2013 was a bit more successful, as they finished with a 74-88 record and a 4.44 team ERA. Still it wasn’t good enough to contend for a playoff spot and definitely not good enough to compete for a World Series title. In fact, in 2007 when the Rockies had their only World Series appearance, they carried a team ERA of 4.32. Only four teams since 2007, including the Rockies, have had a team ERA of over 4.00 and made it to the Fall Classic. The others were the 2009 New York Yankees and Philadelphia Phillies and the 2010 Texas Rangers. As the Mets and Royals have shown us this year, quality starting and relief pitching can take you pretty far in this game. My question is, with all the different strategies the Rockies have tried, what can they do differently to compete?

My suggestion is a slight tweak on an idea that Dave Fleming wrote about in 2009 called the 3-3-3 Rotation. In his article he describes the 3-3-3 Rotation as three pitchers, pitching three innings, every third day with a pitch limit of 40-60 pitches. By having a pitcher essentially go through the order one time, it allows them to give it all they have for a short time instead of conserving their energy for the later innings. In theory, this makes sense. Look at the Royals the past few years; they’ve turned a number of former starting pitchers into relievers and most, if not all have found success in their new roles. In his first year as a starter in 2008, Luke Hochevar had an opponents batting average of .243/.289/.319 the 1st, 2nd and 3rd time through the order. His last year as a starter in 2012 was a little bit better with a .288/.263/.294 BAA but his best season in the majors came as a reliever in 2013 when he held opponents to a .169 BAA.

This may hit a little too close to home for Rockies fans but last year Franklin Morales as a starter for Colorado had a split of .300/.337/.220; in his first year with the Royals out of the pen he held opponents to a .246 BAA. Staying with the Kansas City bullpen, we can look at Wade Davis, who actually had declining BAA numbers in his last year as a starter — .280/.251/.236 — but still posted a solid .151 BAA in his first season in relief. Andrew Miller had a split of .336/.261/.300 in his last year as a starter in 2011; his first year as a reliever in 2012 was significantly better with a .194 BAA. Zach Britton is a similar case with a .272/.266/.293 BAA split in his last year as a starter in 2011 and a .180 BAA in 2012 as a bullpen piece. The point is, generally speaking, when a major-league hitter has a chance to see a pitcher three times in one game, the advantage shifts to the hitter, and if a pitcher with quality stuff can face the order once, the advantage goes to the pitcher. This point is even more important for the Rockies who can’t afford to give their opponents any more advantages when playing in Colorado.

The Rockies have always struggled to attract top-tier starting pitching, since no one really wants to inflate their numbers by pitching half of their games at Coors Field. Colorado has tried to draft and develop power arms who rely on strikeouts and ground balls more so than fly-ball pitchers but still the results are the same ;; a sub-.500 team with an ERA over 5.00, which is not a recipe for success. The average major-league team has five starting pitchers and carries seven relievers in their bullpen. My tweak on Dave Fleming’s 3-3-3 rotation would be to split the 12 pitchers into four groups of three, all with a pitch count of 40-60 depending on effectiveness. In a perfect world every pitcher would go through the order once, throwing anywhere from 30-60 pitches and then turning the ball over to the next guy up who would hopefully do the same thing.

But we don’t live in a perfect world so by having four groups of three, each pitcher could be shifted around depending on the amount of pitches thrown in a week, meaning an effective pitcher could pitch as much as three to four times a week. The average starting pitcher in the majors pitches once maybe twice a week, each time throwing anywhere between 70-120+ pitches depending on the outing; by splitting up that workload they could see action three to four times a week. The average reliever definitely pitches less innings, around 70-80, and in turn throws less pitches but many major-league relievers spent time in the minors as starters, throwing 100+ innings a season. The workload is definitely something to monitor but in 2015 the Rockies used 29 different pitchers. The average amount of innings that a team played was 1,447, and the Rockies staff as a whole pitched 1,426.1 innings. So between the 29 different pitchers, you could keep arms fresh and put pitchers in a position to succeed.

Which brings me to my next point — putting pitchers in a position to succeed. When an offense has a strong 3 and 4 hitter, a manager may put a young player in the 2nd spot instead of lower in the order to ensure that the young player will see strikes. A pitcher never wants to walk someone in front of a player who can crush it out of the park. This leads to more balls seen in the strike zone, hopefully leading towards a positive result, Josh Donaldson is a great example of that this year. Joe Maddon has also implemented a strategy to set young Addison Russell up for success by having him bat 9th after the pitcher instead of 8th before the pitcher. The logic is the same — Russell will see more strikes because opposing pitchers don’t want to walk him and turn the lineup over to their heavy hitters.

I believe the 3-3-3 rotation does this for pitchers, especially pitchers in Colorado. The Rockies had a collective split of .298/.339/.351 the 1st, 2nd and 3rd time through the order in 2015. By having their pitchers face the opposing lineup once, it allows them to display all of the pitches right away. Instead of establishing your A and B pitches the first time through the order and showing your C and possibly D pitches through the second and third time, a pitcher can show all of them through the first three innings. This creates confusion for the hitters and also forces them to be more aggressive at the plate early, something that can be taken advantage of if properly executed. It’s also worth mentioning that some of the best offenses in the game do a tremendous job of communicating with their teammates about the pitcher and the pitches they’re seeing. Remember, the more familiar the pitcher is to the batter, the more advantage the batter has. If you can remove that advantage from the opposing offense, it further sets your pitching staff up for success. Opposing teams would have to have different game plans for each pitcher they see, and those quick adjustments aren’t the easiest to make throughout a 162-game season.

All in all it’s an experiment and besides Tony LaRussa trying something similar for a week in 1993, there hasn’t been another team to try this method. For many teams, the classic five-man rotation works and who am I to say they’re wrong but the Rockies have never really been able to figure it out and if any team is in a position to give it a shot, I believe it’s them.



Print This Post

newest oldest most voted
tz
Guest
tz

You’d have to be careful to have a “spare” pitcher or two available each game in case one of the scheduled pitchers is getting clobbered, and you might want to designate one of your top pitchers to finish off a game where you’re tied or have a small lead.

But, besides the TTO advantage you mention, you could also tweak the approach for home games by scheduling your best-hitting pitchers to be the second pitcher each game. This way, you can have them enter the game as a pinch-hitter for the starter, and then leave the game after they’ve batted a second time, thus improving your offense at Coors without having to burn a position player as a pinch hitter.

All I know is that the Rockies need to do something differently about their pitching. This could be a good step in that direction.

Buddy Bell
Guest
Buddy Bell

tz, this would be a HUGE step in the direction of doing something differently. Can you imagine a close game vs. the Giants at Coors Field in September where the Rox run out Jon Gray for 1-3, Lyles for 4-6, Matzek for 7-8 and have Otto close things down in the 9th! That would be awesome!

Unfortunately, IF Dick “trust the organization” Monfort ever OK’d this, and Weiss had the balls to actually follow through it would really look something more like this:
*In his third start of the season, Gray has a rough 1st inning, giving up 3 hits and a walk for 3 early runs but ultimately settles down and finishes the 3rd inning with 5 strikeouts and 60 pitches but faces 15 batters (Gray actually pitches well in his second time facing the first 6 batters but is pulled due to the 3-3-3 system);
*After a pinch-hitter for Gray in the bottom of the 3rd, Lyles comes on and pitches a solid, but nothing special 3 innings, giving up a 2-run blast in the 5th and another in the 6th. Due to pitch-count restrictions, Lyles makes way for Matzek…
*Top of the 7th and the Rox are down 4-6 (still close) until Matzek implodes, allowing a 5 run inning for the Giants…to stop the bleeding Weiss calls for Kahnle to get the last out of the 7th. Kahnle pitches well enough to get through the 8th unscathed
*After a double-switch in the bottom of 8th (Rox still down 7-11) Weiss gets Flande to eat the last inning so as not to worsen the pitch-count disaster this game was for the 3-3-3 system. Flande gives up a pair of hits and throws 26 pitches to finish the game.

Rox lose 7-11, Gray throws 60 pitches, Lyles 56, Matzek 37, Kahnle 30, and Flande, 26. Rather than using 3 pitchers, the Rox use 5 and push Kahnle towards one of several mid-season injuries due to over-use in mop-up time. Weiss gives up the 3-3-3 system by June 1st and we go back to the typical 5-man, 5+ era rotation we are used to seeing.

Scott
Guest
Scott

A thought just occurred to me:

Have the Rockies REALLY ever brought in an Ace? Hampton was big bucks and piled up W’s on good teams before he came to Colorado but his peripherals were not those of a dominator.

Ubaldo Jimenez put up seasons that make one believe a true “Ace” can still put up relatively dominant numbers despite Coors.

Sadly, young aces are generally not for sale on the trade market and they haven’t been linked to any real franchise #1 SP on the FA market since getting burned by Hampton.

Unless a few more Ubaldo Jimenez types come through the system we may never know what a starting pitcher with really great bat-missing power stuff will do in Coors. Here’s hoping Gray is the next guy who can consistently dominate Coors. (Hat-tip to Jhoulys Chacin for managing several years of competent #2/3 SP numbers w/ COL).

What do we think Kershaw or Jose Fernandez or Scherzer would do playing for the Rockies? Is there a level of skill/mix of elite skills potent enough to nullify or greatly lessen the Coors factor?

tz
Guest
tz

If a sample size of 5 starts is ok, then there are actually some FA options that would fit the bill.

But if you want to see what a stud pitcher might do at Coors, just ignore the numbers for Maddux. And Schilling. And Hudson. And even Hideo Nomo, whose ERA was pulled down by him throwing the only no-hitter ever at Coors:

(Caution: not for the squeamish. Viewer discretion advised.

http://www.baseball-reference.com/play-index/split_finder.cgi?type=p#gotresults&type=p&as=result_pitcher&offset=0&match=career&min_year_game=1914&max_year_game=2015&min_season=1&max_season=-1&min_age=0&max_age=99&isActive=either&throws=any&split_1=locat%3Asite&sid_situa%3Aleado=situa%3Aleado%3Aany&sid_situa%3Atkswg=situa%3Atkswg%3Aany&sid_oppon%3Aoppon=oppon%3Aoppon%3Aany&sid_situa%3Atimes=situa%3Atimes%3Aany&sid_plato%3Aplato=plato%3Aplato%3Aany&sid_role%3Asprel=role%3Asprel%3Aany&sid_situa%3Abases=situa%3Abases%3Aany&sid_outco%3Aoutco=outco%3Aoutco%3Aany&sid_hitty%3Atraj=hitty%3Atraj%3Aany&sid_lineu%3Alineu=lineu%3Alineu%3Aany&sid_situa%3Apitco=situa%3Apitco%3Aany&sid_total%3Atotal=total%3Atotal%3Aany&sid_dates%3Ahalf=dates%3Ahalf%3Aany&sid_locat%3Astad=locat%3Astad%3Aany&sid_situa%3Ainnng=situa%3Ainnng%3Aany&sid_locat%3Asite=locat%3Asite%3ADEN02&sid_dates%3Amonth=dates%3Amonth%3Aany&sid_situa%3Adefpo=situa%3Adefpo%3Aany&sid_situa%3Adr=situa%3Adr%3Aany&sid_rs%3Ars=rs%3Ars%3Aany&sid_situa%3Acount=situa%3Acount%3Aany&sid_situa%3Aouts=situa%3Aouts%3Aany&sid_wpa%3Alever=wpa%3Alever%3Aany&sid_locat%3Ahmvis=locat%3Ahmvis%3Aany&sid_hitty%3Ahitlo=hitty%3Ahitlo%3Aany&sid_situa%3Aclutc=situa%3Aclutc%3Aany&exclude_incomplete=1&c0criteria=&c0gtlt=eq&c0val=0&number_matched=1&orderby=earned_run_avg&sr_split_totals_choice=by_split&order_by_asc=1&sr_pitching_splits_output=view_all&c1criteria=earned_run_avg&c1gtlt=gt&c1val=1&c2criteria=GS&c2gtlt=gt&c2val=5&c3criteria=&c3gtlt=eq&c3val=0&c4criteria=&c4gtlt=eq&c4val=0&c5criteria=&c5gtlt=eq&c5val=1.0&c6criteria=&location=pob&locationMatch=is&pob=&pod=&pcanada=&pusa=&ajax=1&submitter=1

BenRevereDoesSteroids
Guest
BenRevereDoesSteroids

This would be for some fun discussions between players and management come arbitration time.

tz
Guest
tz

Which would be a huge practical hangup for this master plan.

You’d basically have to:

(a) trade the good young arms away for the most you can get (Gray for Schwarber, anyone?)

(b) let any veterans like de la Rosa walk on their FA year

(c) with your pitching staff at rock bottom, sign a bunch of pitchers for dirt cheap who would have fighting to be the 6th or 7th bullpen guy on most clubs, letting them know that they would be part of a 3-3-3 experiment.

Of course, you’d also have to make sure your manager and the organization as a whole were on board with this as well. But once you’ve done all the above, you might be able to establish 3-3-3 as the organizational philosophy, and develop pitchers with this in mind (pitchers whose agents wouldn’t raise a stink ;) )

PandaKOST
Guest
PandaKOST

I would like to see a plan where 9 pitchers throw 1 inning each and pitch 3 out of every 4 games. Again, you break your 12 man staff into 4 groups of 3. I think this helps mitigate the implosion issue, but opens you up to problems with extra inning games.

Game 1: Three person groups A, B, C pitch while D rests
Game 2: ABD, C rests
Game 3: ACD, B rests
Game 4: BCD, A rests

Repeat from the top. Obviously mix and match innings to your strengths. No reason to throw a crumby hitting pitcher if the pitchers’ spot is due in the lineup, don’t throw a lefty with a bunch of righty hitters due up, etc.

James
Member
James

It sounds like you’d also like to see 12 Tommy John surgeries. That would basically be like asking 12 RP’s to pitch 120 innings a year. They would need more than one day of rest after continuously pitching 3 games in row.

Barney Coolio
Guest
Barney Coolio

This idea is interesting, it reminds me of hockey where they have 3 player “lines” or “shifts” or whatever they are called. One issue is that if you always have one guy throw the first three innings, he is shut out from collecting any “wins” for himself. So, even if he has a spectacular year, his record will be something like 0-6, while the middle guy picks up all of the wins. Come to think of it, this could be the best chance baseball has of ever seeing another 30 win pitcher.

Barney Coolio
Guest
Barney Coolio

Suppose this is a spectacular success. Then you basically have 4 starters, 4 middle relievers, and 4 closers. The closers are more like old school firemen who throw 3 innings. And remember, if a reliever enters a game with a lead, and pitches 3 innings without relinquishing the lead, he gets a save regardless of the score. So, under this plan, being a starter is utterly thankles, while being a middle reliever is wonderful. Under a very successful execution of this plan, you could have:

Starter: 0-7
Middle Reliever: 24-9
Closer: 1-1 20 saves.

With no pitcher qualifying for the ERA title, and thus, few Cy Young votes.

tz
Guest
tz

I’d imagine the middle relievers would all be lobbying hard to follow the best of the starters lol.

One thing you could do is have the starters and middle relievers flip-flop roles every other turn in the “rotation”. That might help out if the pitchers are too stat-conscious.

Eric
Guest
Eric

I’ll add these two nuggets and then bid my adieu.
1) The Rockies and the Twins had the most blowout games in the majors in 2014, with 32 and 30 respectively. I defined it as BOTH teams (Rockies and opposition) give up 5+ runs or more. I did not look into 2015 yet. Furthermore, I created a stat that asked the hypothetical question, IF an MLB team could win all 162 of its games by a single run for each contest, how many runs would they need to go 162 and 0? The answer for the Rockies in 2014 was 940, by far the most in the majors by about 100 runs or so. Tells you a lot about the pitching environment there.
2) The typical MLB game averages about 4-4.25 batters an inning so you are NOT going through the lineup 1 time by pitching 3 innings.