Why do many people prefer roll-high to roll-under?

[Jaeger]

That's what I meant about it being a different beast. Once you're using a tabular read it doesn't really matter which way you construct it; the chart is doing the work either way. It tends to make the benefits of roll-low disappear anyway.

Agreed, when you put it all on a table, roll low or roll high doesn't matter.

I do. There's an intuitive quality to roll-low percentile systems that just isn't present with roll high. The handling for most people is different too.

On first blush maybe.

But from a design point of view Roll high does have advantages.

i.e. For contested rolls, you do not need to insert resistance tables, or counter intuitive blackjack mechanics.

High-roll = good, remains a constant.

 

[Thomas Shey]

On first blush maybe.

But from a design point of view Roll high does have advantages.

i.e. For contested rolls, you do not need to insert resistance tables, or counter intuitive blackjack mechanics.

You don't have to necessarily have those with roll low either; you just have to compare gaps (i.e. one character succeeds by 30% and another by 40%). They did that in RQ for years. Its arguably less easy than what you can do with roll high, but whether that's a fair tradeoff depends on how often its used; traditional RQ only used (for skills; opposed attributes were a different story but only one of those came up regularly) it typically for stealth/perception contests which in most games were not a constant thing.

High-roll = good, remains a constant.

Yes, but as I noted, its also notably more painful for people if you also want to use percentile dice (and you may, since you can bake more things into a roll than you can with a D10 or a D20).

(Of course I've reached the point in my life where I'm not sold a big single linear resolution is a great idea whether its percentile or a D20. I'd much rather have 3D6 or a die pool).

 

[billd91]

On first blush maybe.

But from a design point of view Roll high does have advantages.

i.e. For contested rolls, you do not need to insert resistance tables, or counter intuitive blackjack mechanics.

High-roll = good, remains a constant.

It’s not that hard. The current edition of Call of Cthulhu (7e) has a pretty easy algorithm. Each stat and skill gets written on the character sheet with levels of success - normal (full score), hard (half score), and extreme (1/5 score). For contests, the winner is the who rolled the higher degree of success, if rolling the same degree then the PC with the higher score wins the contest.

 

[Thomas Shey]

As an aside, when discussing resisted tasks, since I was thinking of 3D6 systems, there are two at least middlin' well known examples that have a relatively easy way of dealing with that; one roll uses its threshold (how much you made the roll by) as a penalty to the other. You do have to have some method of deciding which is the primary roll, but that's not been difficult in the two cases I'm thinking of.

 

[Aldarc]

But from a design point of view Roll high does have advantages.

i.e. For contested rolls, you do not need to insert resistance tables, or counter intuitive blackjack mechanics.

High-roll = good, remains a constant.

In Dragonbane, when doing opposed or contested rolls, lowest roll wins. Low-roll = good, remains a constant.

 

[Werthead]

I remember when Alternity rolled out with its always roll-under mechanic. It seemed to evolve from one of the 2E complaints, that you were rolling low one second (ability checks) and then high the next (for attacks) but trying to get as low an AC as possible but as high EXP as possible. It felt like a bit of a mess.

One solution was to switch everything to "always roll low" which sounded great, but fans seemed to soon find the loss of "20 to crit, 1 to fumble" to be a bridge too far and there was a lot of complaining about it. I think WotC's research as they took over (they inherited the game from TSR but IIRC it launched on their watch) was that people were not keen on it.

Incidentally it was interesting that Alternity basically had the Advantage/Disadvantage system backed in before 3E even launched, but they used a variant where the second die was a "control die" that added or subtracted an entire die depending on skill level and difficulty. One of those ideas that sounds great until you've done it a few times and then you're trying to roll as low as possible on 1d20 but adding 1d4 on top, which is actually making your score worse rather than better. It's like THAC0 in that it's absolutely simple as a concept, but the tiny amount of extra cognition it takes can be irritating, especially if gaming after a work day and you're looking to do less brainwork.

I suspect the Alternity experience was why D&D 3E launched with "always roll high!" as the alternative mechanic. Which worked out well with the exception of ability scores, which since 3E have been vestigial and useless. Alternity at least made the 1-20 scale for ability scores make sense in itself.

 

[aramis erak]

Agreed, when you put it all on a table, roll low or roll high doesn't matter.

It does when you get to modifiy the roll, and the needed math thereof.

It's like a table in 1d100 range but with all breakpoints at 5 point increments is no different than a d20, unless you have modifiers below 5 points. Which MSH/RMSH/AMSH does. Also, not all the entries are on 5 point breaks; in MSH, it's 5 point blocks to 90, then 91-94, 95-97, 98-99, and 00.

One can totally d20-ize Star Frontiers Alpha Dawn and lose almost no granularity outside attributes... but AMSH, a point of Karma matters at least 1/5 of the time... and 4 points matters at least 4/5 of the time.

In re other Roll-High successful games.
Also, Rolemaster has been a long time stable fanbase, along with companion/derivatives Spacemaster and HARP. Plus the OOP MERP. All of them are 1d100+score, roll high, with 4 basic methods of reading the rolls...

1. 1d100 ↕ + skill total + difficulty +situation mods, <0 is a botch, >100 is a success, >200 is essentially a crit
2. 1d100 ↑ + skill total + difficulty +situation mods, read result from the attack or spell attack table.
3. 1d100 ↕ + skill total + difficulty + situation mods, read result from either the static maneuver or moving maneuver table.
4. 1d100 ↕ + skill total + difficulty + situation mods, final total is percentage of stated action accomplished.

The arrows indicate open-ending directions ↑ is uppper; on 96-00, roll another 1d100 and add, with recursion↕ is double open; if the first d100 is 96-100, as ↑, but on 01-05, roll 1d100↑ and subtract...

There's also a pseudoclone of SM, and a pseudoclone of MERP (Against the Darkmaster) None of them (neither the clones nor SM/RM/MERP/Cyberspace/HARP) are runaway successes, but they have a solid following. a 43 year publishing history with 4+ editions of RM, 3 or more of SM, 2 of MERP (before the license was lost), and then HARP... So, for a single core system across 7 games and a total of a dozen editions.... that's a success, IMO.

 

[Committed Hero]

I suspect the Alternity experience was why D&D 3E launched with "always roll high!" as the alternative mechanic. Which worked out well with the exception of ability scores, which since 3E have been vestigial and useless. Alternity at least made the 1-20 scale for ability scores make sense in itself.

This is how it worked in White Wolf games, that for a time included Ars Magica, which Jonathan Tweet also helped design.

The issue with THAC0 is that it makes little mechanical sense to use a number in the middle of a scale as the reference point. It made sense to current players, because that's how AC functioned for 25 years.

 

[Reynard]

I remember when Alternity rolled out with its always roll-under mechanic. It seemed to evolve from one of the 2E complaints, that you were rolling low one second (ability checks) and then high the next (for attacks) but trying to get as low an AC as possible but as high EXP as possible. It felt like a bit of a mess.

One solution was to switch everything to "always roll low" which sounded great, but fans seemed to soon find the loss of "20 to crit, 1 to fumble" to be a bridge too far and there was a lot of complaining about it. I think WotC's research as they took over (they inherited the game from TSR but IIRC it launched on their watch) was that people were not keen on it.

Incidentally it was interesting that Alternity basically had the Advantage/Disadvantage system backed in before 3E even launched, but they used a variant where the second die was a "control die" that added or subtracted an entire die depending on skill level and difficulty. One of those ideas that sounds great until you've done it a few times and then you're trying to roll as low as possible on 1d20 but adding 1d4 on top, which is actually making your score worse rather than better. It's like THAC0 in that it's absolutely simple as a concept, but the tiny amount of extra cognition it takes can be irritating, especially if gaming after a work day and you're looking to do less brainwork.

I suspect the Alternity experience was why D&D 3E launched with "always roll high!" as the alternative mechanic. Which worked out well with the exception of ability scores, which since 3E have been vestigial and useless. Alternity at least made the 1-20 scale for ability scores make sense in itself.

So much wonderful design and world building alike were lost with the end of TSR Alternity.

 

[Staffan]

Incidentally it was interesting that Alternity basically had the Advantage/Disadvantage system backed in before 3E even launched, but they used a variant where the second die was a "control die" that added or subtracted an entire die depending on skill level and difficulty. One of those ideas that sounds great until you've done it a few times and then you're trying to roll as low as possible on 1d20 but adding 1d4 on top, which is actually making your score worse rather than better. It's like THAC0 in that it's absolutely simple as a concept, but the tiny amount of extra cognition it takes can be irritating, especially if gaming after a work day and you're looking to do less brainwork.

Alternity was a little weird. It was fairly counter-intuitive that a plus modifier made things more difficult. Most roll-under systems add modifiers to the skill value (so if you normally have a skill of 14 on a d20 and you get a +2, you want to roll 16 or lower), but Alternity added its modifier to the die itself. This, together with the degree of success model (roll half your skill or less for a Good success, or 1/4 of your skill or less for an Amazing success) meant that the chances of special rolls were altered much more than the chances of a success itself (if your skill is 14, which is pretty high but not exceptional for a starting character, rolling d20+d6 if you have a +2 modifier does hurt some when it comes to regular success, but it makes it almost impossible to get an Amazing success by rolling 3 or less). In addition, the difficulties scaled a bit weirdly: going from 0 to +/- d4 was pretty chonky, then going to d6 and d8 were fairly small increments, but then it accellerated to d12, d20, and then multiple d20s).

Alternity should, however, be lauded for its absolute dismissal of the d10 as the abomination it is. I don't think it's used anywhere in the whole system.

I suspect the Alternity experience was why D&D 3E launched with "always roll high!" as the alternative mechanic. Which worked out well with the exception of ability scores, which since 3E have been vestigial and useless. Alternity at least made the 1-20 scale for ability scores make sense in itself.

It seems likely that Alternity served as a testbed for some ideas they had for a third edition (even though it was designed under TSR and not WOTC). Alternity stats were 4-14 though for humans, not 1-20.