Just to elaborate on Timz's "Ignore the "starting tractive effort" given for diesels in books (or whatever)." At low speeds, the limit on the tractive effort a locomotive can exert is wheel-slip: try to pull harder than the "adhesion limit" and the wheels spin uselessly. This is why lower-horsepower locomotives are often used for low-speed service: at the relevant speeds, the higher horsepower locomotive would not be able to use all its engine power. (The SD-38/SD-40 comparison Timz alludes to is a good example. As long as a reasonable speed can be maintained, a 3,000 hp SD-40 could do more work than a 2,000 hp SD-38. So the SD-40 was a much better seller than the SD-38. Why did some railroads by a few SD-38? Because they were looking at use at very slow speeds! The U.S.Steel railroads got some: they ran ore and coal trains at low speed. And bigger railroads got a small number for such uses as pushing freight cars over the hump in classification yards.)
Now, about the "adhesion limit". A locomotive moves a train by using the friction -- adhesion -- between its wheels and the rail. With the control systems used on diesel locomotives before the end of the 1970s (EMD's GP-50 and GE's B36-7 introduced more sophisticated wheel-slip control), under good conditions (dry, non-greasy, rail etc etc etc) a locomotive could exert a tractive effort about 1/4 of the weight on its driving wheels: 25% adhesion. (In practice, 25% was optimistic. Prudent railroads typically assumed something closer to 20%, and assigned locomotives to trains accordingly: you don't want to tie up the main line because the locomotives can't get a heavy freight train up the ruling grade! … And modern, AC-motored, locomotives can do significantly better than 25%: over 30%, over 40% in optimal conditions.)
So, for the first few decades of dieselization on American railroads, it became the convention to quote, as a locomotive's starting tractive effort, 25% of its weight-on-drivers. If you have books that give both weights and t.e. figures, check the arithmetic: starting t.e. was almost always quoted as exactly 25% of the nominal weight on drivers! My guess is that railroad officials who had grown up in the steam age EXPECTED to see a figure for starting t.e. on the locomotive manufacturer's specification, and that the manufacturers complied by quoting something easy to compute.