Ricky Romero vs J.A. Happ - an interesting debate

There's some interesting discourse going on right now over the last spot in Toronto's rotation.

(Quick aside/disclaimer: This is a good sign if the most spirited spring debate is over your fifth starter. The most likely scenario is neither of these guys are very good in 2013, but I digress.)

Let's summarize both sides of the argument here:

- J.A. Happ filled in admirably, if not spectacularly, as a starter in a lost 2012 season for Toronto. He is having a good spring.

- Ricky Romero had a terrible year in 2012, and he's having a bad spring.

- Those who say Romero shouldn't have a spot in the rotation given to him say Toronto can't afford to "wait" for Romero to work through his issues in a "must-win" season.

- Those who say Romero should stick in the rotation use sample size in both of their arguments. Happ's spring numbers should be taken with a tablespoon of salt, and you can't omit Romero's fantastic numbers from 2009-11 in your judgment.

Let's break this down, by first putting one weak argument to bed:

- The sample size argument is weak from the pro-Romero crowd, and I'll tell you why by listing these career numbers:

Player A: .500 win%, 590 IP, 4.19 ERA, 1.39 WHIP

Player B: .543 win%, 794 IP, 4.09 ERA, 1.39 WHIP

Player A is J.A. Happ.

Player B is Ricky Romero.

My point: If you're going to take the full sample size for Romero, you have to do so for Happ as well, and the numbers tell us these are two very similar pitchers.

Now, let's put Romero's 2012 season into context:

- His ERA was 5.77, which was the highest in Major League Baseball since 2008.

- He walked 105 batters. Over the last 10 seasons, only 2 American League pitchers have walked more (Victor Zambrano, 106 in 2003; Daniel Cabrera, 108 in 207)

- That walk rate led to batters having an absurd .381 OBP against him. To put that number in context, only 11 players in MLB had an OBP that high in 2012.

When you go thru recent history to find pitchers around Romero's age having 5.00+ ERAs and 5+ BB/9, you find names like Matt Clement, and Russ Ortiz followed by many other names who were out of the league soon afterwards.

In summary, he was bad. Legendarily bad. But, there's also no denying he was very good in the prior years. 

So, the challenge becomes this: trying to find comparables that had initial success, then bombed out in a following season. 

Here's what I did:

- About 100 pitchers log 150 IP or more each season. Usually about 10% of these log an ERA+ of 130 or better, while a similar percentage are below 80. (quick primer: ERA+ is something that compares a pitcher's raw ERA with the league average and adjusts for ballpark factors... the higher the better... i.e. 125 ERA+ means you were well above average, 80 means you were bad).

- Ricky Romero had an ERA+ of 146 in 2011 (9th in MLB), and a combined ERA+ of 119 (22nd in MLB) through his first three seasons in the bigs. It stumbled to 74 in 2012.

- I searched over the last 20 years to see which pitchers had an ERA+ of 115 or better over the first 3 seasons of their career (min. 450 IP). To account for potential outlier candidates (have to consider that Romero might be one), I searched for pitchers who put up one season of 130+ and one season of 80 or worse, all before the age of 28.

Here's the list of closest comparables I could find, listed by my (subjective) order of relevance:

5. Cole Hamels: 38-23, 131 ERA+, 1.14 WHIP in 1st 3 seasons; 10-11, 4.32 ERA, 1.29 WHIP in year 4

Romero wasn't quite as good as Hamels was in his first 3 seasons, and Hamels wasn't nearly as bad in year 4. Hamels rebounded nicely in the following 3 seasons, but there are reasons why I don't think this is a great comparable, mainly because Hamels has never had control issues like Romero.

4. Kevin Milwood: 40-18, 128 ERA+, 1.18 WHIP in 1st 3 seasons; 10-13, 99 ERA+, 1.29 WHIP in year 4

Similar to Hamels, Millwood was better than Romero in his first 3 seasons, and not as bad in the fourth. But he's similar to Romero in that his year 3 was far and away his best year. 

3. Dontrelle Willis: 58-39, 121 ERA+, 1.30 WHIP in 1st 4 seasons; 10-15, 84 ERA+, 1.60 WHIP in year 5

This is the absolute worst case scenario for Romero --- someone who bombed out of the league by the time he reached his late 20s. But there are warning signs here: athletic lefty, control issues, 

2. James Shields: 32-24, 113 ERA+, 1.20 WHIP in 1st 3 seasons; 24-27, 89 ERA+, 1.39 WHIP next 2 seasons

The good: Shields has regained his form to win 31 games in 2011/12. The bad: He's a hard-throwing righty with pretty good control (i.e. is he that close of a comp for Romero?)

1. Joel Pineiro: Through his first 93 starts, Pineiro had a .581 win%, 118 ERA+, 1.26 WHIP... Romero was .592, 119 ERA+, 1.30 WHIP... In Pineiro's 4th full season as a starter, he cratered to a 7-11, 75 ERA+, 1.48 WHIP.

The bad news is Pineiro got even worse the following year. The good news is a couple years down the road he found his game again and came back to have a few positive seasons.

////////////////

In summary, there's just not a lot of precedent for someone who was as good as Romero was in 2011, while also being as bad as he was in 2012. None of these provide a really great comp, but taken as a whole, they provide a pretty good picture of what we can expect.

Given what we've read of Romero's work ethic, I think there's a chance he can rebound like Shields did and have a career like Hamels or Millwood. I think Pineiro is the mid-point range of what we can expect from Romero, as someone who sticks around in the bigs but likely doesn't reach the heights of what he accomplished.

But, I'll close with this: Romero's extraordinary control issues are what worries me, and makes me believe there's a decent chance that he could end up like Willis. 

Want insight into one Memphis executive's thoughts on Toronto's players?

You can access it if you have an ESPN insider account. John Hollinger gives his take in roughly 100 words or so on every player in the NBA... here's key excerpts from his (pre-season) thoughts on each of the principals involved in these trade talks:

Jose Calderon: 

Quick, who led the NBA in pure point rating last season? No, it wasn't Steve Nash or Rajon Rondo or any of the other stars you'd expect. 

It was this guy, who continues to run offenses with uncanny precision. Calderon ranked third in the NBA in assist ratio and handed out 10.4 per 40 minutes while posting one of the lowest turnover ratios at his position, a remarkable achievement. Additionally, he's an efficient shooter who ranked eighth at his position in 2-point shooting percentage and hardly ever misses a free throw. 

But the biggest shortcoming is his defense. Calderon was less visibly awful last season than in previous campaigns, but he remains a speed bump for any guard with speed... Any way you slice it, he grades out as a major liability on this end. That's why, in spite of his tremendous efficiency at the point, his days as a starter are likely over.

Ed Davis:

Davis is long and can jump but badly needs to hit the weights, as he was absolutely thrashed physically by most opposing big men. Despite the abuse, he had an excellent rebound rate, ranking ninth among power forwards, and was 11th at the position in blocks. 

This would be fine if he could play the 4 offensively, but he can't. Davis is a 5 in a 4's body, as he can only finish around the rim and can't create anything for himself. Over his two pro seasons, he's made 31.5 percent from beyond 10 feet, which doesn't cut it.

He is only 23 and still has time to improve, but right now he's a "4.5" who doesn't quite fit at either frontcourt slot.

Terrence Ross:

Ross is a decent perimeter shooter who should improve, and defensively, his length and mobility should allow him to overcome a strength deficit. 

But I just don't see how he gets beyond serving as a role player. He's neither a particularly instinctive nor clever scorer, and he doesn't have a great handle. While he's a good athlete, he doesn't jump off the page. His rebound rate was strong for a wing, but his blocks and steals didn't stand out, and his ballhandling numbers were weak. He struck me more as a mid-to-late first-rounder than an elite talent.

And, we can also get a sense for Hollinger's thoughts towards Rudy Gay:

He can get his shot off against anyone, and there's value in that, but there's not a lot of evidence that he's an above-average shooter... What does have value is when he takes shots near the basket. Gay shot 64 percent at the rim and had a high foul rate for a jump shooter; in particular, if he gets a step going right, he can throw down some vicious dunks...

However, he regressed in other areas. Gay is guilty of dribble blindness, and while he's improved from a few years ago, he still ranked among the bottom 10 small forwards in pure point rating. Defensively, he's had a hard time converting elite athleticism into decent results. His length nets him a lot of blocks and steals, but the Grizzlies gave up 6.0 more points per 100 possessions when he was on the floor.

My take: 

First off, I think it's pretty clear that Hollinger doesn't think Gay is a max-player.

On the Raps front, I don't think anything that's happened over the course of 30+ games this season would cause Hollinger to change his opinion of Calderon, Ross, or Gay. 

Davis, meanwhile, has evolved a bit offensively, hitting 47% of his field goals beyond 10 feet (28-59) this season. Though, he still remains a liability on the defensive end (despite people in the mainstream media who think grabbing rebounds implies solid defence).

However, the Grizzlies don't need bigs. Their starters (Marc Gasol, Zach Randolph) are solid, and they're getting positive contributions from the bench (Marrese Speights, Darrell Arthur) as well.

So, I guess what I'm saying is, if the Raps want to trade for Gay, I think they'd have to involve a third team that could provide a starter at the 3 spot for Memphis.

Who would I turn to? I'd ask the Milwaukee Bucks if they were interested in moving either Ilyasova or Mike Dunleavy (depending on who Memphis liked better).

You can easily create scenarios where Toronto winds up with Rudy Gay, Memphis winds up with Calderon & a Milwaukee SF, and Milwaukee ends up with Terrence Ross and another piece.

And to get you semi-excited, Hollinger's profiles on ESPN of Ilyasova and Dunleavy are both very complimentary.

Shove as many refs as you want...

Amir, you deserve as much shoving as you want, as you're the best Toronto Raptor of the season (which, admittedly, isn't saying much).

Let's break it down as simply as possible:

With Johnson on the court, the Raptors have outscored their opponents by 32 points. With him on the bench, they've been outscored by 160 points.

Why is that? Well, Toronto's rebounding gets much, much better with Johnson on the court. Offensive rebounding percentage jumps 8 percent, while defensive rebounding goes up by 3. The Raptors also attempt more free throws with him on the court (likely a by-product of the offensive rebounding jump), and turn the ball over less often (11 per 48 mins, compared to 14 per 48 with him on the bench).

What does it all mean from a league-wide perspective? 188 players have been on the court at least 40% of the time this season. Here are the 9 players among that group with a net plus/minus above 17:

1. Thaddeus Young 25.2
2. Kevin Durant  24.2
3. Wesley Matthews 22.4
4. Kobe Bryant 21.5
5. Ron Artest 19.7
6. John Salmons 17.9
7. Jared Dudley 17.5
8. Kevin Garnett 17.2
9. Amir Johnson 17.1

Net plus/minus is, by no means, a perfect statistic. It tends to underrate players who play on teams that have good benches, for example. Limitations aside, it still does a pretty good job of showing who's playing well, especially on a team-by-team basis.

Using data from NBA.com, it's not clear whether the Raptors are really rewarding Johnson for his good play this year. There are six possible two-man combinations of Toronto's four most-used post players and I've listed them in terms of net rating (pts per 100 poss - pts against per 100 poss):

1. Johnson-Bargnani 1.8
2. Johnson-Davis 0.6
3. Johnson-Valanciunas -1.5
4. Bargnani-Valanciunas -8.2
5. Bargnani-Davis -9.3
6. Valanciunas-Davis -16.8

Notice a trend with that top 3?

What I find most interesting is that you don't really notice anything about Johnson's personal statistics to explain why the team is playing better with him on the floor. His field goal percentage is down, his scoring and rebounding per minute ratios are right around career norms.

Regardless, the real disappointing thing to consider about these numbers is this: What would Toronto's season look like if they had chosen to bring Valanciunas along slowly, cherry picking only the best possible scenarios for him to play in (10 mins a night), and giving those extra 12 or 13 minutes to Johnson?

The Myths of Point Guard Defence

Defence remains one of the most difficult aspects of basketball to analyze (this, I understand, is a very vanilla statement).  When Player X goes by Player Y along the baseline and finishes a lay up, who's fault is it? Maybe the defensive strategy for Player Y's team was force Player X baseline all game to force him into help defence, and Y's teammate missed the assignment.

ANYWAYS, despite those admitted limitations of analyzing defence, 82games still tries to quantify defence in one way: Opposing PER. The theory is that they track what position each player is playing on the floor, and comes up with a PER number for whoever that player is guarding at the time.

Again, there are drawbacks to this data: It doesn't take into account quality of competition. It doesn't take into account how your team chooses to help you defensively. And, it is still just a measure of efficiency.

However, I'll still take some sort of quantifiable number over the (biased and uninformed) opinions of talking heads.

First, let's build this stat up a bit: Last year's NBA leader in Opposing PER was Andre Iguodala. The top 10 also featured players such as Shane Battier, LeBron James, Metta World Peace, and Luol Deng. All are known as great defenders (and all are small forwards... is that a crappy position offensively in the NBA right now?). And the season before, the top 5 were Iguodala, Dwight

The two best rated point guards were Rajon Rondo and Derrick Rose. Again, this seems to pass the smell test (note: players must have played a minimum of 40% of his team's minutes to be on this leaderboard). On the other end of the point guard scale were Kemba Walker, Jarrett Jack, and Kyrie Irving.

So, I went through every team's page on 82games, and found 43 players who have a relatively significant amount of playing time under their belts, and whose primary position (based on 82games data) is point guard. Here's what I found:

• 7 PGs had a opposing PER above 18: Luke Ridnour, Mario Chalmers, Jimmer Fredette, Kyrie Irving, Jerryd Bayless, Kyle Lowry, and Mo Williams
• 8 PGs had a opposing PER below 12: Chris Paul, Ty Lawson, Rodrique Beaubois, Eric Bledsoe, Eric Maynor, Devin Harris, J.J. Barea, and Norris Cole

On average, the players these point guards were covering averaged 19 points per 48 minutes, putting up a .467 effective field goal percentage, with an average PER of 14.

Which brings me to the real reason I chose to look at these numbers: Here are Calderon's opponents this season: .471 eFG%, 18 pts per 48, 13.6 PER. 

In other words, almost a perfectly average point guard defensively.

And, there's also this: A couple days ago, John Hollinger tweeted the following: "Lowry stopped guarding last year. Big reason he clashed with McHale."

You already saw Lowry on that above list, well, his opponents also have a .549 effective field goal percentage, 5th worst among the 43 point guards we are looking at.

Anyways, the conventional wisdom is that Lowry is this great, bulldog point guard, with a "winning attitude." (the words of journalists who don't know what they're talking about), while Calderon is one of the worst defensive point guards of all-time.

As I said before, I have no doubts that Toronto does things to "hide" Calderon defensively so to speak, which help these numbers a bit. But he's also played a significant amount of starter's minutes, and it's not like he's got Dwight Howard and LeBron James playing behind him.

So, don't believe the hype when you hear that Lowry is a great defender or Calderon is horrible.

82games.com 12-13 season is now live....

It's always a great time of the season when 82games.com decides there's finally enough data to post their pages for this season. They can confirm some things I think I've been seeing, disprove others, and generally allow me to avoid doing productive things. So, without further ado, some bullet points about the Raptors season, using 82games data as the foundation:

• Landry Fields was horrible before being shut down with an injury.
• 82games fact: Fields has the lowest PER of any player in the NBA averaging 20 mins or more per game
• Toronto's point guard play has been solid
• 82games fact: Raps point guards have a joint PER of nearly 18 this season, and PG is one of two positions where the club is outplaying the opposition (in terms of PER), with the other being power forward.
• This has been a tough opening schedule for Toronto
• 82games fact: Over half of the Raptors games this season have come against top 10 defensive teams
• Ed Davis has had another relatively poor start to the season, but he has been productive in one (very particular) role
• 82games fact: Davis and Valanciunas have been on the court together for just 37 minutes this season, but the Raps are +11 in that time. Consider: Toronto's best 2-man combo (Bargnani & Johnson) are +22, and have played together for 144 minutes.
• It's fun to rip on Andrea Bargnani, but unfortunately, it's just factually incorrect to do so --- His 2-for-19 shooting performance yesterday, coupled with a rare sign of life from Ed Davis, had the Bargnani-haters out in full force on twitter. It's great to point out that Bargnani's shooting hasn't been great this season, and his defence is subpar (both of those facts are true). However, the leap everyone takes from that to "He is worthless"/"They should trade him"/"Jonas is a better NBA player" are just wrong. 
• 82games fact: The Raps get outscored by 8 points per 48 minutes when Bargnani is on the bench. A number that is by far the highest on the team. Dig a little deeper and you see that Toronto's offence per 100 possessions is 10 points worse when Bargnani is on the bench. So, despite his cold shooting, the fact is, his presence on the floor opens everything up for Lowry, Calderon, and DeRozan. When he is replaced by an impotent offensive player like Johnson or Davis, that spacing is non-existent, and it makes everything more difficult for Toronto's perimeter players

Defending DeMar... again

DeMar DeRozan has been more a pleasant surprise for me through these first three games than Kyle Lowry. No, I didn't expect Lowry to shoot this well (and none of you should expect it to keep up either), but the defence and attitude were known commodities.

DeMar, on the other hand, has shown signs he is understanding his limitations as a player, and how to maximize what he does well. I've highlighted three statistical examples (let me just get this out of the way now: yes, I know we are dealing with a (very) small sample size. let's move on):

• Rebounding: DeMar might never be a great instinctual defender, but he is athletic, and he has gotten stronger over the off-season. So, at the very least, there is no reason why he shouldn't be one of the most active rebounding guards in the league. And so far, DeRozan's career-best  rebounding rate is the fourth-best in the NBA among shooting guards (minimum 20 mins/game):

1. Klay Thompson - 11%
2. J.R. Smith - 10%
3. Thabo Sefolosha - 10%
4. DeMar DeRozan - 9%
5. Iguodala/Bryant/Curry - 8%

• Ballhandling: Similar to defensive instincts in basketball, passing is a similar attribute, you either have court vision, or you don't. And while I'm comfortable saying DeRozan will never have even average assist totals, he has made up for that by being very protective of the basketball. His turnover rate of 4% is a career-low, and it's third-best in the NBA among shooting guards with a usage rate above 15%.

• Shot distribution: I've talked about this many times in the past, and it's becoming prevailing wisdom even in the mainstream media now: the long 2-point field goal is the most inefficient shot in basketball. However, that shot was one of DeRozan's favourites last season. There are signs he is learning to alter that shot chart. Last season, around 55% of DeRozan's two-point field goals were outside the paint (with the corresponding 45% inside). Through three games this season, those numbers have flipped (45% outside the paint, 55% inside), and it's resulted in a career-high true shooting percentage (59%).

DeMar DeRozan is a player with obvious limitations in his game. But when he maximizes the strong points of his game, he can be a very good player in this league (and good value at $9 million per season).

HIGHEST SCORING STARTING BACKCOURTS IN NBA

1. Houston (Lin-Harden) - 50.6
2. Toronto (Lowry-DeRozan) - 42.7
3. Oklahoma City (Westbrook-Martin) - 42.0
4. Portland (Lillard-Matthews) - 39.6

Of note, Chris Paul and Jamal Crawford are averaging 44.7 points per game, but Crawford is coming off the bench as L.A.'s sixth man.