Is there a consensus on which is better? A combination of both? I noticed some feats offer ASI and some specific boon. Characters also might get better results with ASI increasing spell DC.? Throw in racial choices and it becomes hard to discern the best approach.

I think a mix is usually best. It depends on how you want the character to play.

Agreed.  The best approach is different for every character.  I've found that I end up picking differently with every PC that I create.

My opinion is that by the numbers, ASI is usually better overall. But, that being said, feats have the potential to make playing far more interesting.

Quote from TexasDevin >>

My opinion is that by the numbers, ASI is usually better overall. But, that being said, feats have the potential to make playing far more interesting.

This factor is probably my biggest gripe with 5e. I understand nixing the "you need x and y and z feat to play a two weapon fighter or else you lose" malarkey that happened in 3rd and 3.5th but I really hate having to choose between good statistics and interesting gameplay like the feat vs asi choice often does. It makes it a lot harder to build, say, a Monk, Barbadian or Paladin that's a bit more distinctive than a typical "I attack" bot. Although "half-feats" which increase an ability score by 1 while also giving some other cool effect are a decent compromise, there shouldn't have to be a compromise between making a build that's viable and a build that's individualized and unique.

I feel like the game should be more balanced around every PC getting 2 feats, one at around 5th and one at around 10th. It'd let you vary up your playstyle and experiment with unique or unexpected playstyles(like being a grappler monk, for example) without significantly nerfing you as punishment for daring to break the mold. This is less of a problem with spellcasters than with martials, of course.

But maybe I'm just an extremist when it comes to customization. Were I writing an RPG, I'd probably just nix the idea of classes altogether and let people pick and choose their 'class features' as they level, maybe with some 'class guides' in the back to serve as a suggestion.

But, some people feel that Variant Humans are overpowered by getting a feat at first level too.  It really isn't black and white which route is more powerful.

Quote from TexasDevin >>

My opinion is that by the numbers, ASI is usually better overall. But, that being said, feats have the potential to make playing far more interesting.

Totally agree. I'm in a large (20+ player game) and many of us mentioned that it'd be cool to take a feat, but it's too hard to pass on +1 to your core stat. So, we decided that at level 4 you get both the ASI and 1 feat. It was very interesting to see the range of what we all chose.

The recommendation I've always read is, max your core stat, do whatever else you want with the remaining feats.

The best feat you can take is the feat that gives you proficiency in an ability you're not proficient in.

You instantly get +2 up to a max of +6.

Try getting that with your ASI.

+1 buff (+2 to the ability score) per feat is pittance compared to that. And most of the feats that give you proficiency into an ability, also have some other trick too.

Constitution is the one you want to have.

Proficiency beats the PANTS off advantage, advantage only works out to +2 per roll.

Quote from IDNeon357 >>

Proficiency beats the PANTS off advantage, advantage only works out to +2 per roll.

Advantage is an average of +3.325. But it depends on the number you need to roll on what benefit it gives. If you need to roll an 11 then the benefit is +5. If you need to roll a 20 then it's not even +1. The thing is how often are you required to need a 20? If you average the required roll for most DC's the benefit is typically somewhere around +4.

So no proficiency doesn't always beat the pants off advantage. Usually it starts to win out at level 9+ when proficiency is +4.

As for the topic itself I feel that ASI's will usually be better unless you've got your eye on a specific feat that rounds out your build. If your main stat is Dexterity it's even less of a choice simply because of the huge benefit a +1 dex modifier has. Extra AC, initiative, hit and damage.

Quote from Emmote >>

Quote from IDNeon357 >>

Proficiency beats the PANTS off advantage, advantage only works out to +2 per roll.

Advantage is an average of +3.325. But it depends on the number you need to roll on what benefit it gives. If you need to roll an 11 then the benefit is +5. If you need to roll a 20 then it's not even +1. The thing is how often are you required to need a 20? If you average the required roll for most DC's the benefit is typically somewhere around +4.

So no proficiency doesn't always beat the pants off advantage. Usually it starts to win out at level 9+ when proficiency is +4.

As for the topic itself I feel that ASI's will usually be better unless you've got your eye on a specific feat that rounds out your build. If your main stat is Dexterity it's even less of a choice simply because of the huge benefit a +1 dex modifier has. Extra AC, initiative, hit and damage.

 Well yes, I agree fully, but it will definitely beat the pants off it after level 9. But also, I disagree with the +3.325. First it would have to round down, but also in a lengthy discussion I once had about "familiars" we settled around +2 because of how advantage drops off as you approach 20.

But +3.325 is the average benefit of advantage. The average for 1d20 is 10.5 and the average for 2d20 drop lowest (advantage) is 13.825 for a benefit of +3.325. That's just the way it is.

That said I know that d20's are rolled to beat DC's so the benefit really depends on the typical natural roll you need as I said before. It's certainly not +2 though. The closest you could get to +2 is an average benefit of +2.125 if you only care about rolling a 17, 18, 19 or 20, but what proportion of DC's in a game are ever that high that require a natural roll of 17+? Not many I'm sure.

Most natural rolls needed fall within about 7-16 range and if you average that the benefit comes to +4.575. If you include up to 20 (7-20) then it's 3.875. Nowhere near +2. So +4 is probably the most accurate answer for the typical benefit advantage will provide.

D&D is a game that is notorious for having convoluted rules that make the knowledge floor high for inexperienced players. Less knowledgeable players have enough to remember about their character before feats which makes Ability Score Increases (ASI) a viable alternative. A feat can reward players who have a good understanding of the game and want to specialize, while an ASI is a simple improvement that is easy to understand and use. For the deeply knowledgeable player, feats can be more useful and exciting than ASI. Therefore, ASI and Feats are equal depending on the player's needs. 

My rule at the table is that your first ASI selection (class level 4) you get both an ASI and a feat.  After that, it is back to one or the other. 

Notice that feats are an OPTIONAL thing, that tells you that the game is actually designed around increasing your stats. This is why many feats give you both a stat increase and an ability (or several) to keep it balanced against the benefits of a stat increase. This is also why benefits without stat increases generally have the biggest benefits so as to make up for the losses incurred by not taking the stat increase. That said, some feats are almost required for some builds/classes/subclasses especially at high level hence the rule to max your key stats early then take feats.

Nahh, feats aren’t required at all. People only believe frats are required because they are an option and people make builds for optimization and feel like not taking those feats is so horribly inferior as an option that they would have to staple the back of their hand to their forehead and flounce around saying “oh woe is me” and “cry lament” or something. 🙄

In truth, while it’s is fun to make builds, the entire concept of planning a PC from 1st to 20th level based on some preconceived notion for a game like D&D is nigh useless (somewhat ridiculous even) outside of AL. For a game like WoW, or any other MMO it makes perfect sense to come up with builds, because one builds against an AI that only has a finite number of exact specific options, monsters, missions, environments and strategies  available that it has been programmed to have. Therefore the AI will always be a known/predictable opponent, so a zeitgeist of builds most optimized to overcome that one, predictable opponent is a useful thing. If all eight billion humans on earth played WoW, we would all be guaranteed to have a very similar (if not practically identical) experiences.

For D&D, the closest one will ever find to that is the AL. Outside of that one, standardized format, if all eight billion humans on earth played D&D, we would all be all but guaranteed to each have a fairly unique experience compared to everyone else. For one thing, everyone does/thinks/perceives things in a slightly different way. So while the AI for WoW will always do what it does the same for most people, every D&D table has its own unique supercomputer interpreting D&D’s gazillions of  rulesfrom. On top of that, every DM has their own personal houserules and/or homebrews which cannot be accounted for in any build  but those unique to each table. (Heck, my regular Wednesday evening table has different houserules depending on which one of us is DM for that campaign.) that doesn’t even begin to account for variations between settings, adventure, or even game style. My regular group is a perfect example:

1. Our main GM uses his own custom homebrewed campaign setting inspired by a favorite IP of his; runs a game that’s got rails under at least 50% of the story; and doesn’t use prewritten adventurers so he writes a new chapter to the narrative every week between sessions
2. I typically use a customized version of Mystara as my setting; my games have as few rails involved as I can possibly get away with; and I use a mixture of homebrewed adventures and the old, old modules that typically only cover 2-3 levels and can be swapped around however I need them to be.
3. Our second alternate DM uses his own custom world, does one of those campaigns where everything is on rails hidden behind the illusion of a sandbox, and likes to find interesting user posted adventures online.
4. And our fourth DM uses FR as her sett and uses the pre published 1-fin campaigns (that have falsely appropriated the term “module” even they there is nothing modular about them whatsoever 🤨).

Our Wednesday table, has traditionally been out main GM’s game backi The Before Times (aka pre-Covid). I had also started running a second regular campaign every other Saturday,. That table had half of the people from the current lineup, and three other folks in that game were all also DMs of their own games. The 4th DM I listed was also rrunning  a game one once/month hat included half of us from the Wednesday game, as well as some friends of hers from work. (Those two campaigns have been put in stasis for two years which means their dead, 😟 “le sigh.”) The DM I listed as our “2nd alternate” only returned to the area from out of state about a year before everything went to Tartarus in a teakettle. He used to DM for a group up in NY, but we were so saturated with games that he opted to be an “Alt DM” for the weeks the current active campaign doesn’t run for whatever reason, and he went West Marches with the West Marches style this time.

Out of a total of eleven(ish) total players across those three campaigns, seven of us are/were/had been DMs. Of those seven, I was one and have been a player in campaigns run by four of the other six. Of the 5 DMs who’s styles I can speak on from first-hand experience, no two of them were even remotely similar to the AL, not even remotely. Any build planned according to the suggestions all over the interwebs would have inevitably have performed differently in each of them than they do in AL.

In addition, I am also a player in a PbP (in which the other 3 players are/were/have been DMs too, and There has been enough shop-talk to convince me that builds created according to those internet guides would have yielded somewhat differing results. On top of that, the actual DM for that game disallowed feats.

All of that being said (typed), the single most significant detail that makes any perceived difference in power levels between using feats or ASIs is the fact that the DM’s job is to juggle smoke and mirrors. It doesn’t matter how powerful the party is, the challenges they must overcome should be adjusted accordingly. If the PCs are all optimized to 1337+ combat monsters, if they can barely hit the broadside of a Tarrasque with their peashooters, or anywhere in between; the challenges they will end up being presented should have been balanced against that particular party. That’s why I keep stating that the only “balance” I really care about is the one at the table itself regarding the players ideas about how to D&D. I can help rebalance their PCs a li’l if need be, and I already have to balance my encounter difficulties anyway. As long as the fog machines have dry ice and the mirrors have been polished to a streak-free shine, the rest is covered.

It's about options and feeling unique, both in character and in mechanics.  There is nothing wrong with wanting to thematically synergize mechanics.  However, eliminating feats may take something away from a large portion of the player base that get a lot of enjoyment from them.  I would hasten to say that a table that plays without feats is most likely in the minority.  The reason isn't always about wanting to be OP.