The Many Deaths of the OSR

This is what I have pieced together about the death of the OSR, or "Old-School Renaissance" (of Dungeons & Dragons) based on statements of those involved.

In 2001, Wizards of the Coast created its Open Games License. Now players could copy and publish old editions of Dungeons & Dragons with impunity. They did. OSRIC (2006) replicated AD&D and Swords & Wizardry (2008) revised and replicated OD&D. Other clones appeared thereafter in great number.

In 2008, Gary Gygax died and the Fourth Edition of D&D appeared. Both were lamented by players of older editions of D&D. Calls to "take back our hobby!" began and talk of an "Old-School movement" grew into assertions. The blog race was on to identify the genetic code of early D&D rules and practices that would resuscitate the Original Way of gaming. For a couple of years, there was a lot of energy and debate about what that would mean.

In 2010, the first OSR blog devoted to slandering other OSR bloggers and OSR game designers began. It is so foul that I'm not giving a link to it. Others followed in spirit. OSR trolls appeared, too, some individuals who have spent years of their lives leaving nasty remarks on OSR blogs. As I looked into the development of the OSR, the same names would come up again and again spewing negativity. This kind of contentiousness slowly came to characterize the OSR movement, as its own participants would describe it (below).

In 2011, bloggers started talking about the death of the OSR and the commercialization of the movement and what that would mean for OSR.

In 2012, talk continued about whether the OSR was dying. There was the sense that it had achieved its goals already. So what would it mean now? More OSR products, as long as they were not from Wizards of the Coast. It was no longer about recovering D&D. It was about DIY gaming under an "old-school" D&D umbrella.

At the end of the year, one of the founders of the OSR, James Maliszewski, stopped writing his Grognardia blog. The knives came out the next year as fellow gamers skewered him for not delivering the megadungeon he promised, despite his personal problems.

In 2013, there was evidence of still more discussion about the death of the OSR.

In 2014, D&D 5th edition came out. Subsequent discussions considered whether 5e is an OSR game, could be used for an OSR game, or was influenced by the OSR scene.

In 2015, the new edition of D&D appeared in renewed discussion about the death of the OSR. This spurred some reactions.

In 2017, signs of heated spats between OSR authors showed up. Also, Frank Mentzer, one of the revered BECMI D&D designers and a crony of Gygax, was accused of sexual harassment and more.

In 2018, the fractiousness continued. For example, one OSR blogger ended his blog because of perceived politicization of the OSR. OSR commentators revealed how correct he was by their mean-spirited responses to his calling it quits. An OSR artist "withdrew" because of "the toxicity of the scene." Wherever one stands on the individuals involved, the point is that the nastiness of the OSR community was recognized by its own participants.

At the beginning of 2019, one blogger observed that the OSR was "fractured, fragmented, and splintered." A month later, one hitherto highly regarded, but contentious, OSR designer was accused of rape. The scandal prompted OSR folks to take sides and to repent of their relationships over the subsequent months.

G+ was a Google-based social network that had been home to much of the OSR interaction for several years. It was also the medium for rifts between OSR participants. When it closed up in March 2019, players scattered to other, mutually antagonistic forums that trash-talked each other. OSR now had political sides riven by bitter mutual imprecations. By the middle of the year, players were talking about "post-OSR" and posing alternatives.

In August, one blogger's advice was to "Kill the OSR." More discussion of the death of the OSR ensued. Was it dead? Or not, not, not, not!

At the end of 2019, one blogger pointed out at length that OSR now meant mutually contradictory things. By including variant meanings, this differed from most other blog posts and messages, recurring since 2008, that attempted to define what was really OSR. OSR had become incoherent ("amorphous"), not only socially, but with respect to its meaning.

In 2020, the man in charge of Judges Guild, one of the "old-school" companies that had designed RPG supplements since the '70s, was acknowledged as racist. Designers had already begun actively to distance themselves from the OSR. For example, Joseph Goodman, of Goodman Games, publisher of Dungeon Crawl Classics, stated that DCC was simply not an OSR game. But DCC included an appendix with a list of OSR blogs to follow for inspiration. This was a change in stance. Another OSR game designer basically said that his game is only OSR by association, and that the OSR has become balkanized and that the label OSR "has started to lose traction."

Internet trolls. Bullying. Allegations of rape. Racism. Political divisions. Spats and name-calling and mutual recriminations. A fractured movement of fantasy gaming that has been declared dead, over and over, for nine years.

I have simply offered links to some of what OSR gamers have said. I could have given more links to all of these unpleasant aspects of the OSR, but these are enough.

This is how the OSR movement looks to a veteran gamer who has returned to playing role-playing games only several months ago. You don't have to search far to find this kind of thing.

What does the OSR have to offer old gamers like me or new gamers like my kids in 2020?

I guess you just had to be there!

The Original D&D Skill Rules

Critics of role-playing games with skill systems sometimes support the idea that the Gamemaster should decide outcomes by fiat or by ad hoc rulings on the odds, be it in negotiation with players or by assessing the quality and likelihood of the player’s description of the character’s action. Though often presented as a return to the sources and a putative original play-style, and away from more complicated “modern” rules, the impulse to go back to the sources is quite modern. It’s also right in line with the most ardently free-form storytelling gamers, only with a strongly empowered GM. Fewer rules, more GM fiat.

Some seem to think that the original D&D rules of 1974 do not include skills. In fact, they do. It’s just that they did not call them skills, just as they did not call it a “role-playing game” in 1974, either.

The original skill rules were disorganized. They have no uniform method of resolution. Some are “have/not have” skills. If you have them, they work, and if you don’t have them, you can’t do that action. Some are resolved by a roll of 1D6, with odds specific to character types.

For example, here are some OD&D 1974 combat skills. Different character classes have these different skills distributed amongst them.

• Universal weapon proficiency
• Proficiency with blunt weapons only
• Proficiency with daggers only
• Deadly accuracy with missiles
• Proficiency with +3 Magic Warhammer
• Universal armor proficiency

The weapon skill levels are not expressed in figures specific to each character, but to each character’s class and level. That’s why they are on a separate chart. But every single OD&D character has a combat skill.

Every OD&D character also has a skill at finding secret doors. These are skills that cannot be improved. You have a chance of 1 in 3, or 2 in 3 for Elves. Likewise, every character has a “hearing through doors” skill at 1 in 6, or 1 in 3 for Elves

Of course, all spell abilities are skills, too. These are more complex.

You get the idea. It is not that OD&D lacks a skill system. It’s just that the skill system is so ad hoc and asymmetrical that it does not look like a system, and it’s a poor one, making provisions mostly only for events in a dungeon full of monsters.

When the earliest players, in the 1970s, cried for more “realism,” one of the things they actually wanted was rules to cover more things that PCs could attempt, to minimize DM fiat and maximize DM impartiality.

My personal tastes are averse to complicated rules and large numbers of stats. In this I’m quite in line with OSR play-styles. But gamers who say they are against skill systems don’t know what they are saying. They certainly use skills in their games, but they restrict them to a minimal and incommensurate set, and simply call them character class abilities or by no name at all.

The first game system to incorporate skills by that name was Empire of the Petal Throne (1975). They were mostly professional abilities, enriching character backgrounds and giving context for Referee rulings all at once.

Fortune Favors the Bold: An alternative to XP for GP

In the old days I would give experience points (XP) to players for good role-playing and, generally, success in adventures, not specifically for gathering treasure or defeating foes. That was in line with the common practice for practically all games besides D&D, a game that gradually diminished in importance across the gamersphere during the years I'd been playing ('80s through the mid-'90s) until I stopped playing for a few decades.

It's worth emphasizing, in 2020, when the attention of gamers seems to be overwhelmingly focused on one version of D&D or another, that most role-playing games--those hundreds of other games--don't give XP for treasure or killing per se. So my way of doing this in the days of old was perfectly normal and mainstream.

Now that I'm back to running role-playing games, and coming to terms with a world full of gamers most of whose attention, if not their entire practical experience, seems to be limited to finely-grained variations in D&D, I've moved even farther away from the D&D rules. I see little reason to use XP at all. The fun of the game is its own reward. Characters just advance gradually as I call it. When a Chapter is done, each character gets a little advance. If it's Fifth Edition, then characters just go up a level after every so many three-hour sessions. A rule of thumb is that you go up a level after a number of sessions equal to your current level. That's not a hard and fast rule. If you play short sessions (90 minutes), double the number of sessions needed to ascend. My kids, who play, don't get bored this way, which is important. If you have older gamers who want a harder slog, simply increase the number of sessions required to gain a level.

But if it's my non-D&D home rules, the light rule-set fashioned by my preferences, then the game is designed to give a little advance to a stat at the end of every session, provided your character did stuff. Even failing can lead to improvements, as in life. But there is slow, gradual, nevertheless palpable, low-fantasy advancement every time you play a three-hour session.

My case for ditching XP entirely, even from a D&D-style game, is that the Referee is already stocking adventures, be they dungeons or whatever, in advance, and in effect is distributing the range of XP available already. Referees already secretly plan how fast PC advancement should be. Just let them advance as you planned and cut the accounting. The players definitely need to come up with adequate motivations for their characters in the adventure at hand rather than XP itself. In my view, this is a matter of player skill.

But many friendly commentators said, in response to my notes, that giving XP for gold pieces (GP) in value recovered creates a fun feeling of treasure-hunting excitement.

Was I missing out?

I decided to try to tie the treasure hunt into game mechanics to incentivize that kind of play to see what happened. I just had to do it in a way that did not tie character ability improvement in wizardry or swordsmanship or languages or some other skill to the number of gold pieces a character found. The reason I could not do it that way is that it makes so little sense that it basically ruins it for me.

The solution I devised, to play around with, is a rule with a name: Fortune Favors the Bold.

In my home game, each player character has a Luck stat. It works basically just the same as the one in Fighting Fantasy, which is a heavily modified version of the original Luck stat from the original Tunnels & Trolls. If you are not familiar with it, see what I wrote here.

In my game, Luck is a finite resource. The more your Luck is tested, the more it runs out. And you can even rely on it voluntarily, hoping it will help you in a pinch. But gradually, as you encounter hazard after hazard, it slips away. Your Luck can run out.

Other games that use this Luck mechanic just restore points after a night of sleep or between adventures. The rule Fortune Favors the Bold says that you don't get any lost Luck points back without meeting goals of the adventure.

I'm running a necessarily modified and re-statted version of Gillespie's Barrowmaze megadungeon to test this out. There is one goal in this adventure: getting treasure. The Fortune Favors the Bold rule here means that each substantial treasure the players find leads to the restoration of one spent point of Luck at the end of the session.

Substantial is a relative term. Right now they're mucking about the parts of the Barrowmaze that have been picked over already by earlier tomb-robbers, so even finding a purse of coins is important to them. This is especially so because I've slashed the quantities of treasure in each find-spot, reflecting the greater purchasing power of coins in my setting as compared with most D&D worlds.

But here's another catch: If the characters don't really try to find the grave goods they set out to get, Fortune herself will turn her back on them: I take a Luck point away, to be restored only when bolder behavior appears--but their survival gets more precarious with a lower Luck score!

Fortuna, in my setting, is the goddess of adventurers. Indeed, she favors the bold. But it does not actually matter here whether Luck is really the favor of a spirit, or a goddess, or a subtle matter of reflexes, or narrative sense, or a collection of various intangibles. The Luck score and the mechanic itself comprise a meta-game trait, so tying meta-game goals like collecting treasure with a meta-game trait works fine for me. At least, I find it more palatable than allowing treasure to bestow power-ups, as it works in traditional D&D.

(Getting treasure for greedy characters is an in-character goal. In that way the motivation to get treasure is not a meta-game trait. But compelling players to drive their characters to get treasure: that is a meta-game concern. And an abstract Luck stat is a concern of the player, not of the character.)

In my game, the Referee must declare at the beginning of the session what kinds of actions count as Bold for the rule to work. If reaching a milestone counts as Bold, then you need to achieve that to restore Luck. If defeating foes is Bold, then likewise. My family, who are testing this rule in a scenario based on a treasure hunt, feel the pressure not to give up in a session that has turned up little treasure. They know that their next excursion into the Barrowmaze may not be so fortunate (because of their reduced Luck scores) if they don't make progress or if they pass up the possibility of finding treasures so as not to risk their skins. It's all about incentivizing risk and making sure the players refuse to allow their characters to come home empty-handed.

Give it a try and let me know how it works for you.

What's that you say? You don't have a Luck stat in your game? Well, you may have saving throws. Try this: every time a character makes a saving throw, whether it's successful or not, then make the next one cumulatively harder by one. At the end of the session, instead of giving power-ups for treasure (XP for GP), allow them to reduce their accumulated saving throw penalties. I guarantee you that if they know this is how it works, they will be strongly motivated to keep exercising their character and take the kinds of risks that we all want to see them take in an adventure game.

Audentes fortuna adiuvat!

Luck Stats in Early Role-Playing Games

It may seem strange for a character to have a Luck stat in a game that already determines outcomes of events through dice that generate random results. Why do you need a luck score? Isn’t luck a matter of dice rolls?

The answer is yes, it is, but Luck stats do interesting and useful things.

To be lucky in any role-playing game, the numbers you need to roll against are determined in advance to move the story through one dilemma to the next through the dice. When you made your character, you may have determined stats via preliminary dice rolls that fix parameters for subsequent dice rolls. Or maybe you made your character by distributing points to stats, hedging your bets on what you want to roll dice for, and maximizing the likelihood that you’ll succeed in doing the things you want your character to do most.

Either way, your odds of success are usually linked to things on your character record sheet that you determined beforehand, which represent your odds of success in different capabilities.

But what about those events in the story of the game that do not clearly depend on your descriptive stats like strength, intelligence, constitution, dexterity, charisma, or your skills, your ability to scale sheer surfaces, your ability to hear noises through old doors, or whatever other explicit and implicit stats you have? What do you roll to avoid an impersonal hazard in which character volition or ability plays no real part? Did a vine happen to be present for you to drag yourself from the quicksand? Did you manage to gasp the air before the billow of poison gas surrounded you? Were you standing on the spot where the chandelier fell? Did you fall backwards onto one of the spikes or between them?

The Referee could just invent odds by fiat and roll. The Referee explains that the trap hits you with a poisoned dart on a 1, 2, or 3 on 1D6. You just wait to see if the Referee clobbers you for going into the adventure according to the odds that were set arbitrarily in advance or invented on the spot. As the player, you are passive. The impersonal bad things either happen to your character or pass by. The player is just a spectator, and the player character is a victim, in these situations when the Referee decides.

Luck stats address this problem. Luck rolls put the outcome of impersonal game hazards in hands of the player, whose dice results determine the results of these kinds of situations: avoid hazard or not? Instead of sitting passively, waiting for a judgment by the Referee who rolls the dice in view or not, or just makes the call to hurt your character or not, the player rolls the dice on his or her own in the hope of avoiding the impersonal threat, or getting a break, out of sheer luck. The player is gambling with fortune, not waiting to be stung by Referee decision. It’s about the feeling of taking a risk oneself, instead of opposition by the Referee who is supposed to be impartial.

Some Referees have the players roll for the occurrence of Wandering Monsters. The transfer of dice to the hands of the players for such events has an effect on the player’s experience of the game similar to that of a Luck stat.

The Luck stat is your good friend when you are the Referee. You do not need to make up arbitrary odds for random or unexpected events when you can just tell players to test their Luck stat, for which the odds are inherent to the character affected. Players whose characters get a lucky break through a Luck stat feel that it is deserved, not Referee favoritism.

Saving Throws as Luck Tests

The original Luck stats were D&D’s saving throw scores in 1974. They were not called Luck, but that’s at least partially what they are, in effect. The only character stats involved in determining saving throw numbers in old D&D are class and level (number of hit dice), the most primitive and fundamental stats in that game. The first saving throws therefore combine skill (class and level) with pure luck in their concept.

Saving throws entered D&D through its wargame roots. It was a mechanic to avoid instant death for important characters by simulating the sorts of heroic breaks that heroes and powerful entities are expected to have in fiction. The fighting man somehow leaps out of the way of a blast of dragon fire. The cleric turns her eyes away just before locking gazes with Medusa. Failing these rolls often meant instant death and leaving the game, at least temporarily, so the stakes were high, but the game’s fun was enhanced by allowing that player to attempt a saving roll to create one last chance to save the day. Otherwise, the Referee just says, “You turn the corner and... whoops, you see a Medusa! You’re petrified. The end.”

As a boy, I was boggled by the saving throw charts in my first D&D book (Basic D&D by Moldvay). The arrangement of the numbers seemed arbitrary, as they still seem to me today, and I wondered what the game’s designers were thinking. Why can my character resist deadly poison on a 15 instead of 13? Why does going up a level make me more immune to poison? Who made these numbers up? Also, the names of the saving throws did not correspond to the things that happened frequently in the game, either. Save versus Death Rays? How often do you encounter death rays?

The only plausible explanation I have seen for D&D saving throws, and the rationale that may have lurked behind them, comes from DM David. You can read it here.

Luck instead of saving throws

The first wave of D&D players, using the 1974 edition, experienced a combination of addictive thrill at the new kind of entertainment mixed with confused dissatisfaction with the shambles of theoriginal  rules and their nearly incoherent presentation.

Immediately, gamers designed new systems meant to turn the clunky and unclear features of D&D into something that functioned better.

Tunnels & Trolls, published in 1975, gets too little credit as an extraordinary early re-envisioning of what simpler and clearer rules could do for dungeon adventures. Among the innovations, Ken St. Andre and friends designed a new kind of saving throw.

The Wisdom score was not replicated, and there were no Clerics, but a new Luck stat was added alongside Strength, Intelligence, etc.

The Luck stat was the basis for the T&T saving throw system, replacing a large and complicated arbitrary chart of numbers versus odd effects with a unified mechanic applying to all characters. In T&T, whenever your character encountered an impersonal hazard, from death rays to traps, you could make a Luck saving throw to avoid or minimize it. The original rule was that the deeper in the dungeon your character delved, the harder the Luck save became, whether it was to avoid a trap or the ill effects of some magic or something else. On the first level of a dungeon, take [20 minus Luck] and roll that number or higher on 2D6. Doubles add and re-roll. If you match or beat the target, you are Lucky. If not, ouch. Every deeper level in the dungeon adds five to the target. For example, on level 2, you need to beat [25 minus Luck]. Traps and hazards get trickier the deeper you go, but stats, including Luck, are raised as your character gains levels of experience (contrary to D&D, in which stats change little if ever). Raising stats is, actually, the main effect of character level gains in T&T, where high stats are the mark of advanced characters and more meaningful than character level itself.

A high Luck score also improves odds in combat, along with high Strength and Dexterity.

This was the first role-playing game to have a Luck stat as such. It worked by simplifying a complex saving throw system into something that could be grasped instantly but that also depended on a mysterious personal characteristic: how favored your character is by fortune. It could be a mixture of quick reflexes and divine favor and who knows what else, but it sure helped to be lucky.

Call of Cthulhu

The excellent investigative horror game Call of Cthulhu (1981) was based on the rules developed for RuneQuest (1978) abstracted as the genre-neutral rule set Basic RolePlaying (1980). CoC included several innovations that had a major influence on subsequent games. The best known of them is the Sanity stat, which had its own mechanics and was derived from the core stat Power (representing one’s magical and psychic strength and willpower), determined by 3D6.

Luck was another secondary stat based on the Power stat. Power x 5 equals starting Sanity, but it also equals Luck. You tested Luck as a percentile score with D100 whenever there was need for arbitration over impersonal events. Did anybody get hit by the falling roof shingles? Did you step on a crumbly surface or solid surface? Roll vs Luck to find out.

I used Luck a fair bit when I ran Call of Cthulhu over several years.

Fighting Fantasy Gamebooks and Advanced Fighting Fantasy

In the UK, Ian Livingston and Steve Jackson (not the American one) published the first Fighting Fantasy Gamebook in 1982, Warlock of Firetop Mountain. Its success led to a long series of fun books on the same pattern. These choose-your-own-adventures with minimal randomly-generated stats for your solo character and dice-rolling for combat were popular and created a large fan base. I was one of the early players in the USA. I loved them, especially Steve Jackson’s wonderful Sorcery! quartet (1983-1985).

The spare FF game mechanics are manifestly inspired by those of Tunnels & Trolls. Combat is a contest of opposed rolls as in T&T, rather than a series of alternating blows as in D&D and most other RPGs. There is also a Luck stat. Luck, in FF, is one of the most important features of your character, because there has to be a way to arbitrate impersonal hazards especially in a game without a Referee.

The brilliant innovation here is that your Luck can run out, like Sanity in CoC. Roll 2D6 against Luck. Instead of 2D6 to match or exceed a number, though, you roll equal to or under the character’s Luck score. Whether you are lucky or not, every time you Test Your Luck, the Luck score drops by one, making the next Luck roll riskier.

In T&T, the deeper you go into the dungeon, the harder Luck saves become. In FF, this is mirrored by a gradual attrition of Luck as the adventure moves along. Take one too many risks, make one too many incautious moves, and fate will catch up with you. But sometimes bravery or beneficent divine forces will restore a Luck point.

The Fighting Fantasy gamebooks led to a spin-off full role-playing game system, beginning with Dungeoneer (1989), the inception of the Advanced Fighting Fantasy system. I never saw these books for sale anywhere in game stores in the USA, although it was apparently popular in the UK. The FF Luck system is completely intact in AFF. The main change is that, with continuing characters, Luck is restored back to the full amount between adventures, and experience can even raise your Luck score.

When I started running role-playing games for my kids in the last three years, the second system we used (after Hero Kids) was Advanced Fighting Fantasy in the second edition by Graham Bottley (2011). I admit that I am quite dissatisfied by the poor editing, sloppy presentation, and cost of the AFF2e books and its supplements for what you get. It’s a shame, because the system is great at its core. My kids and I had fun with it, and it inspired me to design my ruleset, which I am play-testing with them now.

Little did I know then that there was another spin-off of AFF already published: Troika! (2017). Here we see the same system reconfigured for weird fantasy in the vein of Terry Gilliam films or “gonzo” “OSR” science-fantasy adventures. It includes the same Luck stat mechanics.

What T&T and AFF have in common (besides a 2D6 Luck roll based on a core character stat by that name) is the sense of pushing your Luck as you go. In T&T this manifests as increasingly difficult Luck Saves the deeper you go. In AFF it’s literally a diminishing resource as your Luck score drains away.

Imagine that after every time you made a saving throw in old editions of D&D, the next saving throw was at a cumulative -1 until the end of the adventure. This might mirror the wonderful effect of rising suspense that the FF Luck system generates.

In Fighting Fantasy, players can voluntarily Test their Luck. Success gives an edge in a round of combat (just as Luck gives an edge in T&T combat, but by a different mechanic). Players can deliberately spend Luck and use it up in this way. Some other games soon adopted the spendable Luck points as a mechanic.

As I mentioned, Call of Cthulhu used a static Luck score, although I have read that by the Seventh Edition of CoC (2014), percentile Luck points can be spent to improve chances of success on other rolls, and then slowly earned back again. Luck thus becomes a diminishing resource if you rely on it, as in Fighting Fantasy.

Luck as meta-game credit to get out of fatal trouble

Other games adopted the mechanic whereby luck is effectively a meta-game credit that players have. That is, they drop the idea of a Luck stat against which you roll the dice, but they keep something like Luck as points you can spend. For example, in Warhammer Fantasy Role-Playing (1986), at least in the first edition (the only one I know well), characters have Fate Points. You only have a few, but you can spend one to alter a dire negative outcome and save your character’s life. Fate Points slowly accrue upon great successes in the game, and you spend them only when in desperate need.

In this way, WFRP’s Fate Points resemble the use of voluntary Tests of Luck in Fighting Fantasy, though they are each much more powerful than a Test of Luck in the latter.

I am sure that other games have copied this, but I got rid of most of my game books long ago and I don’t remember other examples.

EDIT 7/25: I just remembered that the James Bond 007 role-playing game of 1983 introduced Hero Points, which are somewhat like Fate Points. The idea was to permit characters to get away with cinematic feats. You burn them up to assist in success, and you are rewarded Hero Points for successful play. I'm sure that other games had features like this, too. If you know of one from before 1983, let me know.

Luck recharging in cycles of real time

GURPS (1^st ed. 1986) includes a Luck trait that you can buy through the point system, GM permitting. It lets you roll three times for one outcome and take the best result. The meta-game nature of the trait is explicit in that the player can use this once per hour of game play (not of in-game time). Later editions allow this to be scaled up to “super” levels for games in which that is appropriate, so that it can be used more frequently in real time with a bigger character point investment (every half hour, every ten minutes). Unlike WFRP's Fate Points or other versions of meta-game credits, this is a personal trait that is expended but returns automatically in real time.

Overview and back to D&D

Luck traits evolved out of wargaming saving throws to avoid instant death. Luck as a stat has been around in role-playing games since 1975, practically the beginning of the hobby. It puts the action of play related to impersonal hazards into the player’s hands, engaging players in the random determination of their character’s fate, and it frees the Referee from making some arbitrary judgment calls. Often it is used as a diminishing resource, and various means exist for replenishing Luck stats such as finishing an adventure, reward for success, in-game divine boon, or passage of real-world play time.

D&D never developed a real Luck stat. Its players have tended to want to preserve its core features while minimizing innovation, and the saving throw system evolved in a different direction. In the Fifth Edition, you can have “proficiency in saving throws” granted by your character class, based on certain core stats. This means that core character stat bonuses are added to your saving throws. Basing saving throws on potentially any of the core character stats is an old innovation, as in other early games like The Fantasy Trip (1977-1980, with its ST, DX, and IQ saving throws) and later early editions of Tunnels & Trolls (which allow saves versus Dexterity, Strength, Constitution, etc., as well as Luck). Inspiration dice in 5e also can act like weak versions of Luck points in other games, giving a character a one-time expendable meta-game advantage as a reward for good play or the like. For players who are wedded to D&D or never ventured far from it, what may seem innovative (or corrupted) in later editions of D&D, by comparison with early D&D editions, is often really just material imported from other early role-playing games.

[EDIT: I followed this with discussion about another use of Luck points here.]