Bench-Pressing: Character Creation by the Numbers (2022)

People with goals succeed because they know where they’re going.

~Earl Nightingale

Players new and old often ask for help in creating or maintaining their characters. Often those that they have suffer from debilitating vulnerabilities, or simply cannotsucceed at their intended strategies as consistently as they would hope. The objective of this guide is to give general purpose advice for character creation that is not specific to any given build type or class.

This guide is not intended to be an ‘optimization guide’, which go in-depth on character options to accomplish specific goals. It is instead a ‘viability guide’, which only provides general purpose insight to create characters that can survive and excel in their adventuring career.I’m now going to qualify what I mean by ‘viable’.

(A) The character should be good (neither merely passable, nor necessarily excellent) in their primary combat tactic.
(B) The character should have no defensive measure that fails more than half of the time.
(C) The character should have a secondary combat strategy at a passable level of efficacy for when their primary will be ineffectual (ie, immunity to mind-affecting vs. an enchanter character, swarm traits vs. weapon damage characters) and/or a method of ensuring that their primary strategy works even in those unusual circumstances (such as a swarmbane clasp).
(D) The character should serve some function for the party when not in combat, preferably more than one.

Any character that meets these criteria is at least a viable character. Whether they are ‘good’, ‘optimal’, ‘broken’, etc. is beyond the scope of this guide. This guide is simply meant to ensure that whatever character you build, you will know that they are unlikely to die outright, and will be useful to the party. Going above and beyond thecriteria abovecan be done in a great variety of ways, and exploring these options is both one of the great joys of tabletop gaming systems, and entirely beyond what I can discuss in one blog post.

We will nowdive into benchmarks for offensive tactics, then defensive tactics, and a brief glance at skill allocation. I will wrap up with some quick step-by-step examples of a character building process.


Iwill try to avoid terms like ‘optimal’, and ‘bad’, as they mean different things to different people. Instead I will use a color scheme. The benchmarks we will be using are directly adapted (read: stolen) from theaverage monster statistics by CR table. This will provide realistic expectations to match with any character. All of these ratings are dependent on comparisons between the statistics of whatever character you benchmark, and the average monster statistics for a creature with a challenge rating equal to the benchmarked character’slevel. You know what, lets just make an acronym for brevity’s sake, AMCREL, for Average Monster: Challenge Rating Equals Level.

Derklord, from the paizo forums, has graciously madea table that derives benchmarks from the Average Monster Statistics by CR table.It is a thing of beauty and grace, and makes this whole process a lot faster.

Bluedenotes those things that are as good as they reasonably could beagainst an AMCREL. This is as good as it gets, and might be considered ‘optimal’.Attack rolls are blue if they can only fail on a Natural 1 against AMCREL’s AC.Saves are blue if your characteronly fails against the AMCREL’saverage primary ability DC on a Natural 1. AC is blue if the AMCREL can only hit your character on a Natural 20 with its low attack, and damage is blue if the character’s Expected Damage Value (or EDV, a derived statistic we’ll discuss later) for a full-attack is equal to the AMCREL’s 50% of average HP. In other words, this character and an equally powerfulpartnerwould 1-round an AMCREL.

Save DCs for offensive casters are a bit different, as realistically it is impossible to get a save DC high enough that an AMCREL fails a save when they roll a 19 in their strong save. Also, casters are generally hitting multiple targets with their effects, so even a 60% save rate will disable a decent ratioof enemies within a 20′ radius or so. A good Blue benchmark for Save DCswould be that they must roll a 17 or higher on their weak save to succeed.

Any given statistic could be taken to even greater heights, but this is really the highest category of benchmark we should consider. Having a Blue rated statistic will not always bepossible until high level, and really should not be the goal of character building. One would generally have to pull some pretty serious shenanigans and min/max to achieve a blue rating, at least until level 10 or so. If you do have a Blue rated combat statistic, you might consider re-allocating resources to get some of your other stats higher. Generally well-balanced characters are more satisfying to play, and the character will be more useful overall.It would be unreasonableto expect anycharacter to possess a multitude of Blue rated statistics.

Greendenotes those things with a 70-90% chance of a favorable outcome for our character. The math I discussed above remains valid here, except that instead of an AMCREL requiring a Natural 20 to succeed against you or a Natural 1 for you to fail against them, the AMCREL should require a 15 or above to succeed against you with its low attacks, a 13 with its low saves, and you should require a 7or above to succeed against them. Green damage would be 25% of the creatures HP in a full attack, ie it takes this character and an equally powerful partner 2 rounds to take down their AMCREL.

Every character should really have at least one combat tactic that is at least Green rated. If immunity to that tactic is common (mind-affecting strategies, for example) there should be something else that character can do in combat that is a significant help (summoning or support casting, for example).

Orangeis…passable. To qualify as an orange rated statistic, the AMCRELonly needs to roll an 11 or higher to succeed against you with its low attacks, and you need an 11or higher against them. Save DCs aren’t worth considering here any longer. If the average enemy can succeed against a given spell on anything below a 12, that spell isn’t worth casting. Orange damage is 16.5% of the enemy’s HP in a full attack. (It takes two characters three rounds to take it down.)

This isn’t a distinct weakness of your character, but if your primary offensive strategy is only Orange rated, it could really use work. However, this is what I would call the defensive floor. If you are a character that gets attacked frequently, your AC should not fall below orange. As a caster, you would consider your odds of being hitwhen buffed with effects such as shield, mage armor, and/or mirror image when determining what tier you fall in. If a save requires a 12 or higher to succeed against the AMCREL’sprimary ability DC, you’ll need to strengthen that save until it is at least Orange rated.

A quick note about AC: oneshould consider themselves more with “how many full attacks from a CR appropriate monster can I absorb before death,” rather than AC if they have other defensive measures (miss chances and so forth) to ameliorate incoming damage. Deriving those numbers is simply more complex than I felt was suitable for this guide. Ihave developed a more complete ‘durability’ benchmarking process inThink Tank: How Long Can you Survive?

This is a perfectly good place to be at with a secondary combat strategy. For example, if I have a character that can inflict thefrightenedcondition on enemieswith a Dazzling Display 70% of the time, it’s okay if my EDV is only in the Orange range. I’m only going for damage when my fear doesn’t work, and I’m at least okay enough at damaging baddiesto notwastemy turn. Similarly, a character like a Bard might only be Orange in EDV themselves, but by handing out +2 to attack and damage to the rest of the party, he may increase the overall EDV of the party by enough to balance out his less than stellar martial prowess.

Redyou are bad at. Most people are bad at most things, so this isn’t a concern for offensive tactics unless you actually plan to use that tactic frequently. For example, if a level 5 archer has a +14 to hit with their bow,but only at a +3 with a longsword, his inept swordsmanship is of little concern. He is green rated in his primary tactic, so we aren’t worried.

Where Red becomes an issue is defensive measures. If at level 5 you had a save at +3 or lower, it would mean adjusting things or making purchases to raise it immediately, as you require a 12 or higher to succeed against the average enemy’s primary ability DC. Likewise, melee combatants should not have an AC of below 20 by level 5. If you do, raise it as soon as is possible to be at least Orange rated.

To summarize, a character should have at the leastone offensive ability that is either blue or green rated, a second that is Orange rated or above, and no defensive measure should ever fall below Orange.


Purely baseline stats aren’t necessarily representative of combat reality with your character. For example, if your Warpriest pretty much always uses his first swift action to give himself +3 to attack and damage rolls through divine favor, then those bonuses should be included in your benchmarks.

If you possess strong abilities with limited usages per day (such as challenge, smite, mutagen, bane, etc.) you would want to benchmark both your ‘baseline’ combat metrics as well as when you’re ‘firing on all cylinders’. This will give you a more complete perspective from which to decide if you’re comfortable with your character’s resource allocation.

If Buff/Support is your combat strategy, your benefits are harder to quantify, and somewhat beyond the scope of this guide. Anything you do tonumerically benefityour allies’ stats (such as inspire courage, giving out teamwork feats, haste, etc.)will contribute to combat efficacy…just not really in a way I can easily analyze without knowing your party in particular.

This could really be said of defensive spells as well: by negating enemy abilities or attacks (giving out AC, resist energy, life bubble, what have you) you can dramatically increase your own/ your party’s combat efficacy. I just can’t write that out in raw numbers, because I’d need very specific information about the situation at hand.

These notes tend to beparticularly relevant to ‘gish’ characters (3/4 BAB with 6th level spellcasting), who might be comfortable in Orange baseline combat metrics given their abilities to buff themselves and their allies, and cast troubleshooting spells such assee invisibility, ghostbane dirge, align weapon,etc.

Building Backwards

I thinkthe most common mistake players make when creating a character is creating them at level 1 first. Yes, that is the first level you will likely play at, but it is a poor indicator of what you will be capable of down the line. The first level your character should be built at is the halfway point of the campaign you are going to do. The character should then be built for the maximum level of the campaign. Only after this should the character be built for the beginning of the campaign. This ensures that one always has the tools to build towards their goals, and that one knows which character options should be selected at any given level.

For example, I play Pathfinder Society, which typically only lasts through level 11. I therefore build characters first at level 5. I thenadvance that to where I want it at level 11. Only then do I create the level 1 version that I’ll playin the immediate future.

You don’t need to completely fill out a character sheet for the ‘future versions’ of your new characters, just have an idea what class features, feats, and bigitem purchases you plan to make down the line.

A Build Process Illustrated

Note that while I demonstrate the bench-marking process throughde novo character creation, this guide is really intended to evaluate the performance of existing characters.

Now I’ll demonstrate how to use all of this information together. As my first example, I would like to leverage melee attacks as a martial character. I’ll create a Half-Orc Fighter at level 5 with a 20 point stat-buy as a simple example. His name will be Muffins. Before I begin building, let’s figure out what our benchmarks are. As we are a martial, wewant to know theAMCREL’sAC to determine our Attack roll benchmarks, their Attack bonus to determine our own AC benchmarks, their Ability DC to determine our saves’ benchmarks, and their HP to determine our damage benchmarks.

Here are the average statistics for a CR 5 monster.


HPACAttackAbility DC

For attack rolls our benchmarks are therefore: +16, +11, and +7.

Our EDV benchmarks are 27.5, 13.75, and 9.1. This combat metric incorporates both your attack rolls and your damage output. It can also be more readily adjusted to include AoOs you might take, extra attacks from haste, etc. Basically, don’t sweat your attack roll being low if your EDV still falls where you want it. It’s just a more holistic combat metric overall.*

Our AC benchmarks are 27, 22, and 18.

Our saves benchmark is +13, +8, and +4.

Let’s discuss what EDV actually is for a moment. To know expected damage, we first need to know our average damage on a successful strike. For example, 2d6+11 has an average damage of 18. We’ll label this term K for kill. We now need to know our hit percentage. This is derived from attack bonus and AMCREL AC as discussed above, we’ll just go with 60% (+9 against AMCREL 5). Subtract the critical range (let’s say we’re using a greatsword here, so 10% crit range) and multiply by K to get the amount of damage we get when we hit, but don’t roll in our crit range. We’ll call this our Non-Critical Success. So, 60% hit rate – 10% crit range is a 50% non-critical successvalue. Now we have a 10% remaining, but 60% of that are confirmed crits, and the remaining 40% of 10% are non-confirmed crits. The damage our confirmed crits do is, of course, multiplied by our critical multiplier of x2.

(K xNon-Critical Successes chance) + (K x Non-confirmed crit chance) + (K x Crit Multiplier x Confirmed Crit chance)

So, (18 x .5) + (18 x (0.4*0.1)) + (18 x 2 x (0.6 x 0.1)) = 11.9 = EDV. This is unsurprisingly in the orange, as our decent damage is hurt badly by a 60% hit rate. If our K remained constant, and we simply brought our attack bonus to the bottom of green:

(18 x .6) + (18 x (0.3*0.1)) + (18 x 2 x (0.7x 0.1)) = 13.86.So if you have the minimum green attack roll benchmark, you need about 18 average damage to stay in green at level 5. That assumes a single attack. Multiple attacks will need lower K values to stay competitive if their attack bonus remainsconstant.

As a 5th level fighter we will have 6 feats, two traits, and 10,500 gp in assets to distribute.

I’ll address the defensive floor first, making sureeverythingis at least above our Orange benchmark. Let’s begin with AC, as it is simple. I’ll want Muffins’Dexterity to be at least 12, and have +1 Full-Plate. This only take up 2 points out of my stat-buy (18 points remaining), and leaves Muffinswith 8,000 expected gold remaining (I’m using the Character Advancement chartto determine my expected wealth). Just doing that brought me to an AC of 21.

Muffins’base saves from class levels are +4 Fort, but only a +1 for Reflex and Will saves. That isn’t the easiest thing to fix, but Half-Orcs have a great option here. By taking the Sacred Tattoo alternate racial trait (costing us Orc Ferocity) we get a +1 luck bonus to all of our saves. We can combine this with theFate’s Favored traitto turn that into a +2 luck bonus. We will also buy a Cloak of Resistance +1, bringing our weak saves up to +4, and our Fortitude to +7. We still have 7,000 gold, a trait, 18 of our point buy, and 6 feats remaining after achieving our defensive floor.

Now let’s focus on offense.As Muffins is a fighter, I’ll be spending a fair amount of resources in my ability to swing a sword well.

Muffins’ base attack bonus is +5, and he getsan additional +1 to attack rolls from Fighter Weapon Training for a total of +6. Hewill be using STR to damage, and will therefore have a 21STR at this level (10 points spent in point-buy, +2 racial bonus, +1 at level 4, +2 STR belt for 4,000 gp) which raises himto +11 to hit. I willbuy him a +1 greatsword (2320 gp), and take Weapon Focus (greatsword) to reach a +13 attack roll. He now has 1,000 gp remaining, 8 points for stat-buy, a trait, and five feats remaining.

At this point I will already have parts of my damage factored. The average base damage of a Greatsword is 7, Muffins’21 STR after buyingmy belt grants another 7 damage (5*1.5 for two-handed weapons, rounded down), the +1 enhancement is there, and Fighter Weapon Training adds another 1 damage. This is roughly16 total average damage, which becomes 17.3once we add our chance for a confirmed critical hit. However, Muffinswill miss the average monster on a roll of 1-4, meaning my expected damage value is only 14.08(17.3 * 80% chance of success against AC 18).

I have now hit every benchmark I have set for Muffins, and still have 1,000 gold, 5feats, a trait, and 8 points in hisstat-buy. I’ve decided Muffins has a weak personality, and will drop my Charisma to 7, bringing my point buy total to 12. As he’ll be taking a fair number of hits, I’ll bring his Con to 14, bringing our Fortitude save to +9. I’d like my Will saves shored up a bit, so that will also be a 14 (bringing Muffins’ Will save to +6) and as fighters struggle to get skill ranks I’ll put my last 2 points into Intelligence.

My saves are actually looking pretty decent, so I’ll spend my trait on Defender of the Society to increase my AC to 22.

The 1,000 gold will likely go into a wide variety of knick-knacks and potions, so I’ll just leave that alone.

NowMuffins just has fivefeats remaining. Power attack is an obvious choice. While it puts Muffins attack roll to the bottom of the green benchmark, the +6 damage increases his EDV to 16.94.At level 6, Muffins’ new attack will be a huge boon to his EDV against his CR 6 AMCREL.

I wouldn’t mind getting Muffins’Will save even higher, butReflex saves just aren’t something I care about. I’ll give Muffins Iron Will to increase his Will save to +8. I’ll take Ironhide to increase my AC to 23.My initiative really isn’t very good, so I’ll take Improved Initiative as my final feat. I might even buy a Cracked Dusky Rose Ioun Stone to get to a +6 Initiative modifier.

Still got a feat left over we could put into anything from Weapon Specialization, to a Teamwork feat, to Dirty Fighting, or even Skill Focus (Knowledge: nobility). The benchmarks have been all already been met, so it can go anywhere you want.

By replacing my Darkvision with the Skilled alternate racial trait and using my favored class bonuses for skill ranks, this seemingly simple fighter can take 5 skill ranks each level.This is more than enough to satisfy some out of combat roles, though party face is not going to be one of them.

Muffins now hits an AMCREL 70% of the time, and is well into Green EDV. He only gets hit 40% of the time with high attacks, and 25% of low attacks, and succeeds the average Fort save 75% of the time, Will saves 70%, and Reflex saves 55% of the time. We are offensively capable, defensively pretty good, and skilled enough to be useful outside of combat.

Muffins is by no means an optimized character, but he’s above ‘viable’. I’d probably call him ‘good, but not extraordinary.’ His EDV and attack rolls are green, and all defenses aregreen except for Reflex, which is widely considered the least important save anyway.This is a decent enough front-line fighter. From here I might think about future feat and equipment choices, where to go with class features, and see if I like where it’s headed by level 11. If I do, I’ll finally build it fully at the level it will begin play at.

Now I’ll demonstratebenchmarking a level 5 battlefield control Wizard against the same AMCREL. His name will be Bagels.There are some differences in which statistics from our AMCREL we view as important, however.

CR: 5High Attack: +10Primary Ability DC: 15Good Save: +8Poor Save: +4

Notice first whatBagels isn’t paying attention to. He doesn’t care about enemy AC, because he isn’ttrying to hit with physical attacks. He doesn’tcare about HP, because he isn’tbothering with damage, he’strying to control the battlefield. Hestill needs to know the enemy ability DCs and attack rolls, because he still must meet hisdefensive benchmarks. Hisoffensive benchmarks are now the enemy saves. A Blue save DC by our previously stated standards is 21,and a green is 18. We would never considering having a save DC lower than that in our focused school, as we’ve decided on a very Save vs. X combat strategy. Support/buff casters have a lesser need to invest intheir casting stat. First things first, though. He needs tomeet his defensive minimum benchmarks.

Elf gets us the stats we need as a Wizard, and it’s iconic enough to be in this sort of quick guide. Bagels can therefore spend a mere 5 points from his stat buy to have16 Dex, for a +3 to AC and reflex saves. Betweenmage armor andshieldspells, he hasalready reached our orange AC goal of 21. With some minor illusion spells likemirror image,invisibility,ordisplacement, or any of a wealth of other defensive magical tactics Bagels can improve his own defenses on the fly. In other words, hewon’t need to focus much on improving hisarmor class through feats or equipment, and he’ll try to evade or disable enemies more than confront them head to head anyhow.

Our base saves are only +1, +1, +4. Our base save at least takes us to the Orange in Will, and our Dex gets us there for Reflex saves. Fort saves are a bit trickier, as Elves are CONdumped and Wizards Fortitude saves are weak. That’s a good sign to take a trait like Orphan, which will grant as +1 to Fortitude saves. We’ve spent 15 points from our initial point-buy to start with an 18 INT and 16 DEX, because we want those. We could just put the remaining 5 points in Constitution to raise it to 12 (14 with a -2 racial penalty). By spending 1,000 of our gold on a cloak of resistance +1, we get our final save to the orange benchmark. They are still dangerously low at +4/+5/+4, but we have used few resources and can afford to address our offensive strategy before spending resourcesto reach higher defensivebenchmarks.

So far, Bagels hasonly spent hisbase-stat point buy, 1 trait, and 1,000 gold. We still have a trait, 4 feats (one of which must be a metamagic feat), and 9,500 gold remaining. Wizards are pretty simple with early feats. Our 3rd level save DCs are 17 so far, but when we take Spell Focus (conjuration) and Greater Spell Focus (conjuration), we get 1 past the minimum of Green at DC 19. With the remaining feat, trait, and metamagic bonus feat there are some options now that all of our benchmarks have been met.

The remaining trait could go into a save, as all of Bagels’are pretty weak. It could be Reactionary if we’d rather just go first more frequently. It could be magical knack if we plan to multi-class down the road. It could be Magical Lineageif we want to use a low-level spell with a lesser metamagic cost. Since we’ll be doing a lot of AoE control spells, I’ll take Magical Lineage to reduce metamagic level increases toaqueous orbby 1. I’ll then use my metamagic bonus feat to get Focused Spell, which can now be freely applied to anyaqueous orbBagelscasts. Spell Penetration and Greater Spell Penetration are good, obvious choices for a save-or-x caster, but I feel they are better left for levels 7 or 9 and up, as SR is uncommonly a real concern until that level range.

Hisfinal feat will be more defensive, as I really don’t like having his Fort save this low. I’ll take Great Fortitude to bring my Fort save total to +6. I could instead go for Augment Summons as a more offensive option if I choose. I’ve got some gold left over, maybe take up mycloak of resistanceto +2, and buy an INT headband. Keep it simple. We’re now offensively almost blue rated (our highest level conjuration spells have a save DC of 20), within all of our defensive benchmarks, and have a nifty trick with a particular spell. We will be getting 7 skill ranks per level, as our favored class bonuses are likely going into HP.

Again, not optimized, but it has no truly glaring vulnerabilities, and will be more than capable within its specialty. It also has ample skill ranks to fulfill out-of-combat roles, and with Bagels’ high Intelligence he would make a wonderful knowledge-base, and/or translator for the party.


Once the above has been established, you should have a concrete number of skills to distribute (perhaps making a decision or two where Favored Class Bonuses are concerned) and know what your best stats are. Now it’s time to figure what you’re accomplishing when the party isn’t in initiative order.

Our color-coding benchmarks are a bit different here. I’ll give general rules, but skill DCs vary wildly depending on what skill we’re talking about (Ride DCs are mostly easy, UMD tends to be difficult) as well as what usages you intend to put the skill towards. These benchmarks are therefore far more subject to context and interpretation than those discussed in the combat benchmarking section:

Blueis total modifier >= Character Level + 10. If you have a +15 modifier in a skill at level 5, you’re as good as you need to be at it.

Greenis >= CL + 5, but < CL + 10. This is something you’ve got a good chance of succeeding outright, but failures will happen with some frequency. This will usually be sufficient to use the aid another action.

Below green, the skill isn’t really a ‘role’ for your character, but distributing a skill rank here and there into your ‘non-specialties’ can be wise. Making sure you can at least attempt Acrobatics, Swim, and Climb is generally a good idea. Maybe at least being able to attempt Sleight of Hand and/or Disable Device. Any character can benefit from at least a touch of Perception. It really depends on how many ranks you have to throw around, and what kinds of things you want to be able to at least attempt.

Muffins has a great STR and decent WIS, and should have 25 skill ranks to distribute. Swim and Climb should be invested in, to offset his considerable Armor Check Penalties at the least. Perception, Sense Motive, and Survival are also good bets for him. He might consider gettinglenses of detectiondown the road, and some ropes and pitons to help the party cross obstacles that he can strip down and accomplish readily.

Bagels is more likely to focus on skills like Linguistics, Knowledge, and Spellcraft, but will still likely distribute some into Perception, Escape Artist, and Acrobatics. If I prefer to be a face character as well, I might consider trading my Magical Lineage trait for Clever Wordplay and invest in Diplomacy.


To quickly theory-craft a character that will be both offensively and defensively viable, we complete the following steps:

  1. Identification of the concept, and the basic mechanics (race, class, archetype) you will useto bring that concept to life.
  2. Find the strengths and vulnerabilities of the race and class choices you decide upon, as well as which enemy combat stats you should measure yourself against.
  3. Ameliorate vulnerabilities to bring them within your minimum benchmark by spending character resources such as traits, stat-buy, feats, selectable class features, and gold.
  4. Once defensive stats are brought within your desired benchmark, spend character resources to bring your offensive abilities within your desired benchmark.
  5. Any remaining resources can be spent either offensively or defensively, or upon out of combat roles. In general, once combat benchmarks have been met, skill ranks can be easily distributed to find a suitable out-of-combat role. Depending on the importance you place in your skills, you might consider taking some of your resources out of combat viability and reallocate towards skilldeftness.

Top Articles

Latest Posts

Article information

Author: Arielle Torp

Last Updated: 01/13/2023

Views: 6139

Rating: 4 / 5 (61 voted)

Reviews: 84% of readers found this page helpful

Author information

Name: Arielle Torp

Birthday: 1997-09-20

Address: 87313 Erdman Vista, North Dustinborough, WA 37563

Phone: +97216742823598

Job: Central Technology Officer

Hobby: Taekwondo, Macrame, Foreign language learning, Kite flying, Cooking, Skiing, Computer programming

Introduction: My name is Arielle Torp, I am a comfortable, kind, zealous, lovely, jolly, colorful, adventurous person who loves writing and wants to share my knowledge and understanding with you.