Every time an NPB pitcher transitions to MLB, we see a number of projections for how he might perform statistically. For me these projections are always a bit of a shot in the dark - there is just too much variance among too many factors to make a linear statistical project realistic.
The big, obvious difference for pitchers is simply that MLB hitters are better than NPB hitters. The best MLB hitters are better than the best NPB hitters, and the average MLB hitter is better than the average NPB hitter. But beyond that, there are a number of more subtle factors that make projections difficult. Here are the ones I can think of, from the perspective of NPB:
- There are fewer legitimate power threats in Japan. Every lineup has at least couple of regulars who just don't hit home runs.
- On the other hand, there are rather few strikeout machines like Adam Dunn and Pedro Alvarez.
- Japanese managers are still in the habit of regularly throwing away outs with sacrifice bunts.
- Most Japanese ballparks have massive foul ground.
- The NPB ball is slightly smaller and lighter than the MLB ball. I've also heard that it is a bit tackier than the MLB ball and easier for some pitchers to command.
- Japanese starters, at least the good ones, go a bit deeper into games. 120 pitch starts are on the high end in MLB, but are not uncommon in Japan. Yu Darvish was no stranger to pitch counts in the 140′s in his NPB days.
- NPB starters normally get six days between starts, which includes a weekly off day. Pitchers sometimes get in a rhythme of pitching on the same day each week. The most famous example of this was Lotte ace Choji Murata, who was nicknamed "Sunday Choji".
- NPB starters also stay home the day before their scheduled starts.
- Playing in Japan requires less travel. All of Japan is in one time zone, and five of the 12 NPB teams are based in the vicinity of Tokyo.
- NPB has 12 teams total, six in each league. Normally a starter will face each team in the opposite league once during interlegue play, and see the other five teams in his own league four of five times each. Most MLB starters will see a bigger variesty of lineups.
Stats are good, they just need to get filtered through all this stuff, plus whatever else I didn't think to include. Ultimately the conclusion I've arrived at is that skills are translatable, but stats less so.