Ecth,
Modifiers exceed usual OB not only due to the already discussed mods on the OB and DB tables, but also due to the fact that shield/parry DB is foe specific, meaning that someone else can attack you from a flank or rear negating those (and DB negation = OB bonus). So before you even start applying mods, you have two DBs in play, dependent on position and maneuvering, and your parry can be "avoided" via maneuver, in combat. (i.e. "It sucks when 2 orcs jump me from opposite sides"). . .Simple surprise, which can also "just happen" can toss most of your bonus out the window in round 1. . .missile combat, especially parry, adjudicates completely differently than melee, negating or modifiying DB. . .Bolt or ball attacks may sidestep DB completely. . .an unseen enemy among the foe can enter combat unexpectedly, not as nasty as flatfoot surprise, but still a totally different set of factors. . . .
Combat is not a generic OB/DB vs OB/DB comparrison, it involves pairings of attention as combatants declare on each other, and tactical maneuvering to attack someone who is declared on someone else, from the flank or rear if possible, is a basic element of RM combat resolution. . . .your effective DB of 60 vs their effective OB of 20 might seem good, until the guy you didn't declare on flank attacks you, negating your DB down to 10 and getting a flank bonus to boot.
Morale can have a giant impact. . .a fight to the death vs kill a few and they run. (for either the PCs or the "monsters").
the human factor involved, i.e. tactical sense, has a vast impact on results in RM. . . .
We all know people, who if you put a character in their hands, will proceed in a straight line toward the nearest combat, confront it head on, and engage in no tactics more complicated than OB/DB split.
We also know people who love to play it obliquely, never approach a situation directly, sneak and peek constantly, and are inclined to avoid combat until they can make a surprise attack, or initiate combat from behind cover.
We know people who avoid combat entirely if possible, either via sneakiness, or social skills.
We know casters who love to cast and will use magic in any combat. . . .we know casters who horde PPs like lifeblood and wait until the last possible moment to cast.
It's chaos. . .and you know your players and their characters in fine detail. . . .your specific party is not a generic group of arbitrary professions of identical levels. . .they choose a different mix of weapons than someone else would, they use different tactics, they pick different lists, and they have different equipment. . .every party is as unique as a snowflake, and a good GM knows them in fine detail. You know them to a far better margin of detail than any "generic party" could cover. . . .and your party is not +3 levels or -2 levels from this average concept. . .odds are they vary in power level from the generic in different ways in different situations. You also know the players, and have a feel for their tactical sense, and how they will leverage what they have on paper. . .combat monsters, sneaks, avoiders, sneaky tactical dark knights of doom?
So, in actual play, you have far more input data on your party than is possible to quantify. . .and much of which you probably can't express but know on a gut level.
Then you choose how smart/tactical to play the opposition, an encounter can be vastly stronger or weaker depeneding on how you choose to play the monsters. . .and GMs vary as much as players, some vary the opponants intelligence/tactical sense a lot, some are very straight up combaters, and some are usually very tactical, where the enemy is often quite nasty. The difference, in a game run by me, of how a goblin group will fight, depending on how smart and experienced they and their leader are, is enormous.
It's an art, with a huge chaos factor on one hand, and hard to quantify human factors on the other. . .I can see plenty of advice to offer, but considering how differently things can play out it's really rough. . . .
You run your numbers through a combat simulator X times. . . .does it factor in "orc 1 plows straight in on you, while orc 2 maneuvers to the flank and attacks"? Does it account for "The orcs all execute a fighting retreat through the next doorway, then surround the space on the far side with the spear orcs in the back, so they can execute 6 on 1 attacks on anyone who follows them."?
Those are just 2 examples of maneuvering and tactics that differ from "In a flat featureless plain extending to infinity with no cover, the two groups engage in combat where everyone accepts straight pairing of opponents without any attempts of tactical finesse, maneuvering for effect, use of terrain/cover or anything other than direct face to face combat." . . . .
the CR differs party to party, it shifts with immediate situational effects like surprise, and bounces around round-to-round based on tactical choices. . .the human mind of the GM keeps juggling moving factors in play and works it. . .you don't just mentally go to "OK, this is around a CR 7" and stick there.
That's where it's art, not science, and advice, not a misleadinly simple single number, are more helpful in actual practice.