Measuring the Synergy of Build Features - Feedback Requested

[Orcus Porkus]

(this is a repost from the WotC optimization board where I didn't get any feedback on this. Is this idea really not worth discussing? Please let me know your honest thoughts.)

I've been torturing my head to find ways that allow builds to be measured by other means than DPR and nova capabilities.

I realized that a build can be called optimized not just only when it deals a lot of damage, but also if all class features, feats, powers, paragon path features, epic destiny features, and magic items synergize well. Simple example: Weapon Expertise is synergizing with all attack powers, and any effects that depend on hitting a target.

The total synergy of a build could be measured by cross-referencing every feature of a build to every single other feature, and then count the number of synergies in a build.
Hypothesis: The fewer synergies there are, the less optimized a character probably is. It means, a lot of class features for example are neglected, items serve only one purpose, the build is a one-trick pony, etc.

The image below is an example spreadsheet for my Eladrin Barbarian build. You need to right-click and "show image" on the graphic to see it in full resolution. The ratio at the bottom right describes the relation between the total number of all present synergies and the number of all theoretically possible synergies (meaning, everything synergizes with everything.) This build achieves almost 0.33. So you could say that this build is 33% synergized. It feels very synergized, but right now I have no clue what this ratio really means since I have no comparison with other builds. It's also not complete, and probably has some errors. I didn't put in rages for example. Perhaps one should add a column for everything that matters for all builds, like defenses for example, to be able to better measure features that boost those.

The tricky part is the question, what constitutes a synergy. I realized that anything that has either a direct or indirect benefit with each other should count. For example, Fey Charge synergizes with Fey Step, obviously, but it also synergizes with Swift Charge, Charging Rampage, Rampage, etc. since all those features benefit from the feat.

Another challenge to this idea is the problem that not all features are equally important, obviously. For example, a negligable paragon path feature, like Fleet Skirmisher in this example, bloats up the total of possible synergies, and thus reduces the synergy ratio. But I actually think this is good - it measures that the build takes a paragon path where not all features synergize well, or are completely redundant.

The chart also helps to "rank" features in the build, although you usually have a feel for that anyway. In this particular example it's evident that the rageblood vigor class feature (temp HP on a kill) is not synergized at all. It is an obvious weakness of the build.

If you want to play with the excel sheet and make synergy rankings of your own build, you can get the file here. I would love to hear your feedback, and perhaps other builds put into a chart like this so we can compare and see whether this whole idea makes sense at all.

 

[AbdulAlhazred]

There are just so many things that are subjective. Is a character with only a few narrow choices that work together better or worse than one with with many broad choices that don't all work together.

What constitues 2 things synergizing? You show Eladrin Will synergizing with pretty much everything, why is that? You probably have some logic for saying that, but I equally have to believe that there is good logic to NOT say that. Many things obviously do synergize but many others are judgment calls

 

[Orcus Porkus]

Well I think it's not that subjective actually. With Eladrin Will, I was thinking that +1 on will and +5 on save vs charm effects helps with attack powers working (meaning, you are not stunned etc.)
But that's not an optimal solution. I think defenses should be implemented better in the synergy chart. If something synergizes with will, it can be identified better, and if you do nothing to synergize with will, it objectively harms your build.

I'm actually challenging the notion that builds that focus on a one-trick pony can ever be called truly optimized. In a real game they'll fail, or be boring, or simply banned by houserule because they are so obnoxious. A well-rounded, synergized character on the other hand is more believable, coherent, accepted, and successful in groups.

A character with "only a few narrow choices" can get away by claiming "highest DPR at level 30" or "can kill Orcus in 1 round with a daily nova" but the synergy chart will probably reveal the "hidden" weaknesses - neglected at-wills, un-optimized encounter powers, super low defenses, lack of mobility, functioning only with CA, and so on.

With more and more of the cheesy tricks being nerfed, and stuff like orbizard, salve of power and frostcheese being next, traditional optimization will either learn to focus more on overall synergy or simply wait for the next Dragon article with broken crap released by accident.

 

[Mengu]

If you have a +8 attack, +15 damage bonus on your opportunity attacks, and you have spend 5-6 resources on them, and you get one opportunity attack per 3-4 encounters you're in, there is something wrong with the allocation of resources.

There is also the problem of diminishing returns. I could have 20 healing surges and 300 hit points, but if I never go below 100 hit points, and never use more than 8-9 healing surges, again there is a waste of resources somewhere.

It all really takes an observant player to see what the needs of the character and the party are, and covering those needs at every "level up". If you do this, in my book, you have an optimized character. You don't have to plan 30 levels or theorize how in 27 levels, when all your buddies are around one solo, you're bloodied, the enemy is dazed, the wizard has his zone thingamabob, and your fighter is holding his dagger instead of waraxe, you can use an action point to give the barbarian 10 attacks. This sort of theory is the pitfall of optimization.

 

[AbdulAlhazred]

Except I think a lot of people sling this whole optimization thing around but they don't seem to have really played with a lot of what are considered optimized builds. Actually VERY few of them are these hypothesized one trick pony builds. Most heavily optimized characters actually perform quite well in a variety of situations and in a few, usually quite common, situations they provide some fairly major benefit. Even if a character only gets to do their one really awesome "trick" once every 5 encounters that's fine as they're pulling their weight (IE winning 1/5 of the encounters). As long as they contribute at an adequate level the rest of the time they've earned their keep.

I guess the other real question would be is this matrix really a useful measure of anything? Synergy is often pretty subtle. I think the only way to really measure it would be to play the character in a variety of situations and determine how well each of the character's resources they actually use are contributing to their ability to do their job. You talk about things like "neglected at-wills" but how is it less optimal for a character to not put resources into an option that is virtually always inferior to another option? Focus can make a lot of sense and not all such focus by any means is making some kind of narrow character that can only do one thing. Its just much more complex than that.

 

[Mengu]

Except I think a lot of people sling this whole optimization thing around but they don't seem to have really played with a lot of what are considered optimized builds. Actually VERY few of them are these hypothesized one trick pony builds. Most heavily optimized characters actually perform quite well in a variety of situations and in a few, usually quite common, situations they provide some fairly major benefit.

You're right of course, I wasn't trying to generalize that every single optimized build is a one trick pony. I was just pointing out people sometimes get a bit overzealous with the damage bonuses they can stack on top of each other, and miss the fact that a single feat or item could give them an extra 5 or 10 hit points which could be handier than a once in a blue moon +3 damage bonus. Or they forget part of a striker's job is being mobile, and manage build a character with no mobility and no way of making a ranged attack. Or they fall in a pit and realize they have absolutely no way to get out, but they are the best tank in a pit ever.

I think versatility and tools of an adventurer are underrated in more than a few "optimized" builds. And yes, this stuff can be campaign specific. Maybe a DM runs encounters in 10x10 blank rooms, and the most DPR you can do is the best character ever. But making the decision to build that character should be based on that feature of the campaign, and not just optimization in a vacuum.

 

[Orcus Porkus]

Or they fall in a pit and realize they have absolutely no way to get out, but they are the best tank in a pit ever.

Ah, that's where the term "pit fighter" comes from! LOL

Like Mengu I don't think that all optimized characters are one-trick ponies. I'm only attempting to put a number on builds that are not purely DPR or nova oriented. The chart I put on is really just the first draft. I'm fishing for ideas here.

What excited me when I started to work on it was the fact that it's actually pretty straightforward. You can itemize pretty much everything a character is made of.

And when you take a feat (over another), or a magic item (over another), or a paragon path (over another), or a power (over another) you invest resources, and they either synergize with the others or not. If they synergize in a subtle way, it still counts.

What I really like about the chart though is the fact that it evaluates the character in its own parameters, by cross-referencing its features with each other, rather than with an externally set optimization goal (like DPR, mobility, survivability). Doing this is another step (definitely worth checking out.)

 

[AbdulAlhazred]

Well, presumably you then want to compare with other options or with other PCs.

I think the issue with subtle synergies and interpretation of what constitutes a synergy is really an issue of how comparable two values are. Given the subjective nature of at least some interactions its hard to say how comparable 2 assessments would be. Even assuming they are entirely objective or at least consistent we still don't know how valuable each interaction is. Some things are highly conditional and may well 'synergize' with a lot of other stuff but simply not really matter much since they rarely come up.

 

[Orcus Porkus]

I agree, but I think it levels out, and anybody who would do this chart is already not picking anything that's really weak anyway (feats, powers), except for tons of class features, defenses etc that receive no attention because of the focus on his one-trick pony, for example.
Would you mind trying to put your own favorite level 30 build in such a chart? Give it a try

 

[Mengu]

Maybe some other "trivial" columns could be added to the top row at the end. These would be things like Hit points, Healing Surges, Defenses, Attack bonuses, Damage bonuses, Movement, Skills, and maybe a few others. Because otherwise, it doesn't look like the elven cloak you are wearing synergizes with anything, but in reality it might be stacking with your bonus from race, background, high stat, and training. Also this approach would make it easier to sort of assign a score to how many resources you have spent on each aspect of your character from defense, to healing, to offense, to mobility, to skills.