When people cite BA, OPS, HRs, etc... do they always include games played or PAs?
It's kind of understood that when there are great variances in games played of PAs when comparing two players, we factor in the sample sizes.
We may think player A is better than B based on better numbers per game (or 162 game prorations), but player B is more valuable because he plays way more games.
Why we wouldn't expect people to think the same way with WAR highlights the fact that many don't understand what WAR measures.
Here's a great comp that Sox fans can be in tune with. Let's compare Porcello and Price with the Sox since 2016.
We'd all say Price is the better pitcher and has been the better pitcher over those years, when he pitches. I would not disagree. We all know Price has been fragile, pitches less IP per start and has less starts than Porcello. We all probably agree that IP and GS'd are valuable, too. So, when we think of total value, we have to do sort of mental evaluation on how to weigh effectiveness vs longevity and durability. It's not easy to do in our heads. We each probably value one more than the other in differing degrees.
ERA:
3.85 Price (85 ERA-)
4.28 Porcello (94 ERA-)
xFIP
3.76 Price
4.32 Porcello
WHIP
1.20 Price
1.23 Porcello
K/BB
3.95 Porcello
3.92 Price
W-L
46-24 Price
61-38 Porcello
WAR (a cumulative metric)
10.7 Porcello
10.5 Price
Why? Because of this....
GS
125 Porcello
97 Price
IP
762 Porcello
586 Price
The difference here is 28 starts and 176 IP. That's 28 more starts of 6.1 IP each over 4 years. That's about a full season more of pitching over 4 years. Surely that has major value. How much value is up for debate, and WAR has their formula for counting that. Maybe it is flawed. Maybe our own mental gymnastics are, too.