← Back Published on

The Last Men Standing: Memory and the Record

There came a time when things which should not have been forgotten were forgotten, and when history became legend, and legend became myth.

J.R.R. Tolkien, Prologue to The Fellowship of the Ring

I originally wrote this the week of August 24, after reading Dom’s book. I spent four or five days researching and writing. As best I remember, I shared it with Malaky about a month ago, but otherwise kept it to myself. Eventually, I decided I probably wouldn’t publish it. Then John Romero corrected Sandy Petersen again on October 27, and Rob Kuntz’s recent statements about AI gave me another reason to revisit it. Fortuitously, this discussion continued developing while this article has been sitting unpublished. What follows is the original piece, updated to address those developments.

Something peculiar happens when a man lives long enough to become part of history. At first, he was simply doing the work. He designs games, writes rules, builds dungeons, argues with other designers, misses deadlines, makes mistakes, and occasionally creates something extraordinary. Fifty years later, most of the people who were standing beside him are gone, and suddenly he is no longer simply a participant. He has become a witness, perhaps even one of the last witnesses.

That is where Sandy Petersen and Robert J. Kuntz become fascinating to me. Both men have legitimate accomplishments that almost anyone working in tabletop gaming would be proud to claim. Petersen created Call of Cthulhu, co-designed Ghostbusters, designed levels for Doom and Doom II, worked on Quake, and later contributed to games including Age of Empires. Rob Kuntz was standing near the center of the Lake Geneva gaming circle during the creation of Dungeons & Dragons, became one of its important early players and playtesters, created Castle El Raja Key, became Gary Gygax’s co-Dungeon Master, and co-authored Supplement I: Greyhawk.

These are not insignificant résumés. Conversely, they’re extraordinary ones, which makes what has happened in recent years tough for me to comprehend. Why does either man need more?

Sandy Petersen and the Romero Problem

The Petersen situation is particularly intriguing because another major participant in the history he discusses is still alive and paying attention.

John Romero has repeatedly corrected Petersen’s recollections about the development of Doom and the early history of id Software. It happens often enough that the exchanges have developed a familiar rhythm. Petersen makes a claim about his years at id, and sooner or later Romero appears: “Hi Sandy, hope you’re doing well.”

In one of the clearest examples, Petersen said, “I WROTE the original Doom backstory.” Romero responded: “Hi Sandy. I hope you’re doing well. However, you are incorrect.”

Romero then explained that Tom Hall had written the story before Petersen joined id Software in September 1993, roughly nine months into development. Petersen clarified that he meant the abbreviated backstory used for the game, only for Romero to dispute that as well, saying Kevin Cloud wrote the manual and summarized Hall’s original story. GamesRadar documents the exchange and several others here.⁠

This isn’t an isolated disagreement. In 2021, when Petersen discussed Doom’s sales, Romero responded, “Hi Sandy, hope you’re doing well,” before correcting the figure. When Petersen discussed the origin of rocket jumping, Romero began, “Hi Sandy, hope you’re doing well. This is inaccurate.” In 2024, when Petersen discussed the shotgun, Romero again opened, “Hi Sandy, hope you’re doing well.” The greeting has become almost a calling card for what follows. PC Gamer collects the exchanges across several years here.⁠

The examples continue. In July 2026, Petersen related another story involving a contractor supposedly working on Doom’s dial-up code. Romero responded that Petersen had conflated four different stories and that the contractor episode Petersen was describing actually concerned the SNES port of Wolfenstein 3D and occurred before Petersen joined id. Time Extension preserves Romero’s detailed correction here.⁠

That last example is important because there is a distinction between recalling an event differently and remembering yourself as part of an event that occurred before you worked there. Memory is fallible. We all understand that. People combine conversations, shift dates, confuse firsthand knowledge with something another employee told them, and slowly incorporate stories they heard into stories they experienced. Petersen himself has addressed this. In July 2026, he said that Romero “helpfully chimes in with corrections,” rejected the idea that the two men are feuding, and attributed some of the discrepancies to the passage of time and imperfect communication inside id Software. That deserves to be included because Petersen’s explanation is relevant evidence too. Petersen’s comments appear in the July 3 entry preserved here.⁠

Then, on October 27, Romero corrected him again⁠, this time over Petersen’s account of convincing id Software about deathmatch. Romero pointed to the company’s January 1993 press release announcing multiplayer, months before Petersen joined. His response also raised the possibility that such inaccuracies were intended to generate engagement. That remains Romero’s suspicion, rather than an established motive. Whatever the explanation, the repeated public corrections are documented, and the historical claims can be examined without deciding why Petersen made them.

That discrepancy matters because Petersen isn’t merely telling stories around a gaming table anymore. He is one of the surviving participants in the creation of culturally important games, which means his recollections will inevitably become source material for future historians. Sandy Petersen does not need a larger résumé. If the only major game attached to his name were Call of Cthulhu, he would still deserve an important place in tabletop history. Then there is Ghostbusters, followed by legitimate work on Doom, Doom II, Quake, and later computer games. The documented accomplishments are already extraordinary.

Then There Is Rob Kuntz

Kuntz presents a different version of the same problem because there is absolutely no legitimate history of early Dungeons & Dragons that should erase him. I have written about Kuntz and his contributions to Greyhawk before in Greyhawk: Legacy, and none of what follows is intended to diminish those accomplishments. Quite the opposite. His actual place in the history of the game is significant enough that it doesn't require embellishment.

The original 1974 Dungeons & Dragons rules provide us with an unusually useful historical marker. The title page identifies the game as being by Gary Gygax and Dave Arneson, while separately offering special thanks to the Midwest Military Simulation Association, the Lake Geneva Tactical Studies Association, Rob Kuntz, and Tom Keogh.

That tells us two things simultaneously: Rob Kuntz mattered, and Rob Kuntz was not credited as an author of Dungeons & Dragons. Those facts aren’t contradictory. Kuntz was unequivocally and unquestionably deeply involved in the development of the game and Greyhawk. Robilar became one of the legendary characters of the early campaign. Kuntz created Castle El Raja Key, became Gygax’s co-Dungeon Master, contributed material to the developing Greyhawk campaign, and eventually received an author credit alongside Gygax on Supplement I: Greyhawk. As I discussed in Greyhawk: Legacy⁠, the history of Greyhawk becomes considerably poorer if Kuntz’s contributions are pushed aside.

The problem isn’t giving Rob Kuntz credit. The problem is determining where that credit ends and revisionism begins. Compare the contemporary record with some of the language being used today. Kuntz’s current official biography calls him “one of the legendary founders of the D&D game” and the “Last Designer Standing” from its earliest beginning. The same biography lists him as a primary D&D playtester from the original 1972–1974 playtest team and documents an enormous body of subsequent design and development work. Kuntz’s current Three Line Studio biography can be read here.⁠

That’s precisely why the word founder deserves examination. It is a considerably larger historical claim than saying Kuntz was an important early player, primary playtester, designer, contributor, and Gygax’s co-DM. The original game provides us with a useful anchor precisely because it was published while these events were still happening. The credited authors were Gary Gygax and Dave Arneson. Kuntz was thanked separately. Then, when Kuntz was credited as an author on Greyhawk, his name appeared accordingly. That’s not an insult to Kuntz. It’s actually a remarkably clean historical record of his growing importance.

That’s why I don’t understand the need to push it. Rob Kuntz doesn’t need to become a third creator of Dungeons & Dragons to matter. His tangible accomplishments already guarantee him a place in the history of the hobby.

He already matters.

Kuntz’s recent statements about artificial intelligence introduce a different question about evidence. In an October 24 post⁠, he described three shelved game projects whose underlying system, he says, could not previously be implemented in print or on computers. He now argues that AI can finally make those designs possible, with the technology running the role-playing system itself. His earlier announcement⁠ traces the first project to 1973 and presents its realization as a potential paradigm shift in gaming. As backing, he described a two-day exchange with Grok that he says confirmed the system’s novelty and its dependence on AI. That is a substantial claim, particularly when the confirmation being offered comes from the technology he proposes using.

I admit that this makes me think of Howard Stark in Iron Man 2, explaining that the technology of his time could not realize what he envisioned. What I didn’t expect was for Tony to hand over the campaign and let Ultron run game night. But the actual evidentiary problem is straightforward: a two-day conversation with the technology being proposed as the solution is not independent confirmation of the claims being made for it. Grok agreeing with Kuntz does not establish that his design is unprecedented, that earlier technology could not support it, or that AI is necessary to make it work. Those claims require a working demonstration and independent examination. Kuntz’s accomplishments already deserve recognition; whether this becomes another one depends on what the system can actually do.

Griffith Morgan and the Arneson Corrective

Then there is Griffith Morgan, and Morgan is different from Petersen and Kuntz in an important way. Morgan wasn’t standing in Lake Geneva or the Twin Cities when these events happened. He came afterward. Along with Chris Graves, Morgan created Secrets of Blackmoor, the documentary devoted to Dave Arneson, the Twin Cities gamers, David Wesley, Braunstein, Blackmoor, and the prehistory of Dungeons & Dragons. Morgan has subsequently worked extensively with surviving Blackmoor participants and historical materials, co-authored The Lost Dungeons of Tonisborg, helped preserve original Blackmoor material, and continued that work through Fellowship of the Thing and ArneCon.

His contribution to preserving a part of gaming history that was neglected for far too long is valuable, and there really was something that needed correcting. For decades, popular culture reduced the creation of Dungeons & Dragons largely to Gary Gygax. Dave Arneson could become the other name in small print, if he was mentioned at all. The Blackmoor story, David Wesely’s Braunstein experiments, David Megarry’s Dungeon!, the Twin Cities gaming culture, and the strange series of experiments that eventually produced Blackmoor deserved much more attention than they received. Morgan helped give those people their voices back.

But historical correction creates its own temptation: the overcorrection. In interviews surrounding Secrets of Blackmoor, Morgan has argued strongly against the mythology that Gygax created everything. Speaking to Kotaku, he argued that the Twin Cities group had been working on what became role-playing for years before Gygax became involved and said that Gygax still did not know how to do it until early 1973. The full Kotaku interview can be read here.⁠

There is considerable evidence supporting the importance of those Twin Cities experiments. There is also no serious reason today to deny Dave Arneson’s status as co-creator of Dungeons & Dragons. But something anomalous happens if correcting the Gygax-centric mythology requires us to create an equally simple Arneson-centric mythology, because there is another artifact sitting on the table: Chainmail.

Dominic Martyne and the Evidence Problem

There is also a recent contribution to this argument that deserves attention. Dominic Martyne’s 2026 D&D: An American Institution spends much of its early history reconstructing D&D’s wargaming ancestry and devotes an entire appendix specifically to Rob Kuntz. Full disclosure is appropriate here: Dominic is a friend of mine and a fellow member of a private group of writers and game designers. That friendship does not make his historical assertion correct. The evidence he cites has to do that work.

His book is particularly useful because Martyne does not require us to choose between acknowledging Arneson’s originality and acknowledging the material from which Blackmoor developed. Discussing Kuntz’s argument, Martyne writes that “Arneson’s creation was greater than the sum of its parts, and this is correct.” Blackmoor did become something greater than the individual games and experiments that preceded it. Recognizing that does not require feigning that earlier elements never existed.

Martyne identifies the experimental Twin Cities gaming culture, Diplomacy, Braunstein, Outdoor Survival, and Chainmail among the elements from which Blackmoor emerged. More importantly, he turns back toward Arneson himself. Martyne notes December 1972 correspondence between Arneson and Gygax in which Arneson says he used Chainmail, and he points to the 1977 First Fantasy Campaign as another explicit confirmation. He additionally brings in Dave Megarry and Bob Meyer as supporting witnesses to the use of those rules within the Twin Cities group. Earlier in the book, Martyne reconstructs the famous Bob Meyer troll episode. Meyer played a Hero under the Chainmail fantasy combat rules, failed his attacks against a troll, and was killed in return. According to Martyne’s reconstruction, Arneson subsequently modified the system, adding hit points and further mechanics that would develop toward Blackmoor. That is an important distinction. Arneson changing Chainmail is not evidence that Chainmail was irrelevant. If anything, it is evidence that Arneson was designing: taking an existing system, finding where it failed to produce the experience he wanted, and changing it.

Martyne later describes Arneson replacing the one-hit-kill mechanic, modifying the to-hit matrix, creating additional charts, and continuing to use a 2d6 system. Blackmoor was not simply Chainmail with a dungeon attached to it, but neither does that developmental history support treating Chainmail as though it were merely an incidental book lying somewhere in the room. Adding hit points, altering the to-hit matrix, and retaining 2d6 are evidence of Arneson’s design work on an existing system, not evidence that the system was irrelevant.

This is where I find myself agreeing with Dominic. None of this is necessary. We don’t have to diminish Arneson to establish Gygax. We don’t have to diminish Gygax to establish Arneson. We don’t have to erase Chainmail to establish Blackmoor, and we don’t have to reduce Blackmoor to Chainmail to recognize what Gygax and Perren’s game contributed. The accomplishments are already there.

For anyone interested in Martyne’s work beyond this book, his ongoing writing on early gaming history can be found at Dominic’s Substack⁠. His own essay on Chainmail⁠ likewise argues from Arneson’s statements that the rules formed part of Blackmoor’s mechanical ancestry.

The Long-Vaunted Chainmail Oracle

And somehow, once again, we are back to Chainmail. Gary Gygax and Jeff Perren’s medieval miniatures rules were published in 1971. The fantasy supplement provided rules for heroes, wizards, fantastic creatures, magical weapons, spells, and individualized fantasy combat. Blackmoor unquestionably transformed the experience into something fundamentally different, particularly through persistent characters, dungeon exploration, advancement, campaign continuity, and the role of the referee. But the relationship between Blackmoor and Chainmail cannot simply be wished away because it complicates a cleaner creation story. Martyne’s evidence makes the problem particularly difficult to avoid. If Arneson himself says he used Chainmail, if correspondence from the developmental period says so, if The First Fantasy Campaign says so, and if other Twin Cities participants remember using it, then anyone arguing that Chainmail had no meaningful place in Blackmoor’s ancestry carries a substantial evidentiary obligation.

Perhaps Chainmail really is the long-vaunted oracle after all. Not because everything came from Chainmail. It plainly didn’t. Nor because Dave Arneson simply sat down one evening, opened Chainmail, and started playing D&D. The evidence does not support anything that simplistic either. Chainmail matters because every attempt to create a clean creation myth eventually seems to run headlong into it.

If Dungeons & Dragons was simply Gygax’s creation, Blackmoor is a problem. If Dungeons & Dragons was simply Arneson’s creation, Chainmail becomes a problem. If role-playing appeared fully formed in one Minnesota basement, Braunstein and David Wesely become problems. If everything sprang from Lake Geneva, the Twin Cities become impossible to explain. Perhaps the Chainmail Oracle refuses to die because every time somebody constructs a history requiring Chainmail to disappear, another contemporary document, recollection, rule, or 2d6 table seems to drag the darn thing back onto the table. Maybe those aren’t problems with history. Maybe they’re problems with the way we’ve been trying to tell it.

The Creation Was a Chain

The birth of Dungeons & Dragons increasingly looks less like a light bulb switching on over one man’s head and more like a chain of innovations moving between groups of obsessive gamers. There was Strategos. There were David Wesley’s Braunstein games. There was Dave Arneson’s Blackmoor. There was David Megarry and Dungeon!. There were Gygax and Perren and Chainmail. There was the Chainmail Fantasy Supplement. There was Arneson bringing Blackmoor to Lake Geneva. There was Gygax recognizing that something remarkable had happened and beginning the process that ultimately transformed these experiments into a written, publishable game. There were Kuntz and the Lake Geneva players hammering on the developing rules through play. Eventually there was Dungeons & Dragons.

Remove pieces from that chain and the story becomes cleaner, but it also becomes less complete. That is why Morgan belongs in this discussion even though his position is fundamentally different from Petersen’s or Kuntz’s. Morgan isn’t enlarging his own role in events he personally experienced. He is interpreting the roles of people who can no longer completely speak for themselves, and that brings a different responsibility.

The Gygax-centric history deserved correction, and Morgan and the Blackmoor historians have performed valuable work in correcting it. Martyne himself credits Morgan’s preservation work in the Blackmoor portion of D&D: An American Institution, specifically relying on published materials that have helped make early Arneson documents available for examination. That is worth emphasizing because disagreement over interpretation should not erase useful historical work. But corrective history has to submit itself to exactly the same evidentiary standard as the history it seeks to replace. Documents first. Memories considered carefully. Interpretation identified as interpretation. When the evidence does not allow certainty, say so.

The Temptation of the Last Witness

What we’re watching may be less sinister and considerably more human than people want it to be. These men are getting older, their contemporaries are disappearing, and that inevitably changes the balance of historical authority. When ten people who experienced an event are alive, one person’s recollection exists among nine competing recollections. When only one or two remain, memories begin sounding like testimony, and testimony gradually begins sounding like fact. The last man standing acquires an enormous advantage because fewer people remain who can say, “That’s not exactly how that happened.”

John Romero happens to provide that check for Sandy Petersen. Romero was there, and watching Romero and Petersen disagree demonstrates exactly why historians should remain skeptical of eyewitness recollections, even when those recollections come from people who unquestionably participated in the events. Romero is effectively performing historical peer review in real time.

Kuntz increasingly doesn’t have a Gary Gygax available to do that, nor does he have Dave Arneson. That doesn’t mean Kuntz is wrong whenever he makes a historical claim. Quite the opposite. His firsthand recollections are invaluable because there are things he saw and conversations he heard that were never documented. Morgan has the opposite problem. He has documents and testimony, but he was not personally present. He therefore has to determine how much weight to give competing recollections from people who were.

These are three different historical problems, but they lead to the same question: How do we know what actually happened?

This is where discussions about gaming history often go sideways. People treat proximity as authority. Someone says, “He was there,” as though that ends the discussion. Being there matters enormously. It makes someone’s testimony important, but it doesn’t make that testimony infallible.

Historical evidence has different weights. A rulebook printed in 1974 is extraordinarily powerful evidence about who received authorship credit in 1974. A dated letter written in 1972 is powerful evidence concerning what two designers were saying to each other in 1972. A manuscript, contract, convention program, photograph, playtest document, or contemporaneous journal can all be evaluated within their historical contexts. A person remembering something in 2026 is also evidence, but it is 2026 evidence about a decades-old memory.

Memory changes. People unintentionally absorb stories told by others. Memories combine, chronologies compress, and important later events migrate backward in our minds. People can even remember themselves as being more central to events than contemporary records demonstrate. None of that necessarily requires dishonesty, and that distinction matters enormously here.

This brings me back to the question that started bothering me. Why would men whose authentic accomplishments already place them among gaming’s most significant figures feel any need to expand those accomplishments? Maybe ego plays a role. Ego plays a role in almost everything humans do, but I suspect something deeper may be happening: legacy anxiety.

Imagine watching fifty years of books, documentaries, websites, YouTube channels, museums, and historians gradually determining where your name belongs in the lore. You know what you experienced. You remember conversations nobody recorded and ideas being passed around someone’s kitchen table. Then you open a history book and discover that several years of your life have become three sentences. That has to be strange, and I can understand the instinct to correct the record when you believe the record is wrong.

The danger begins when correcting the record becomes enlarging your place within it. Eventually you aren’t simply fighting over what happened. You’re fighting over what people will remember happened after you’re dead. Participant becomes contributor, contributor becomes designer, designer becomes creator, and creator becomes founder. None of those transitions requires us to determine anyone’s intent. They can happen one adjective at a time, repeated often enough that eventually the adjective becomes part of the accepted history. The same thing can happen in historical interpretation. An overlooked participant becomes important, important becomes essential, and essential becomes the origin. A necessary correction can gradually become another creation myth.

That’s the part I find unfortunate. Rob Kuntz does not need to be the third creator of Dungeons & Dragons. He was Robilar. He was there during the foundational Greyhawk campaign. He created El Raja Key, co-DMed with Gary Gygax, and co-authored Greyhawk, one of the most consequential supplements in the game’s history. That’s already an extraordinary life in gaming.

Dave Arneson doesn’t need Gary Gygax diminished in order for Arneson’s accomplishments to become larger. The historical rehabilitation of Arneson over the last several decades has been necessary and overdue. Blackmoor deserves its place, Braunstein deserves its place, and the Twin Cities gamers deserve their place. So do Chainmail and Gary Gygax.

And Griff Morgan does not need to be wrong about everything for us to disagree with a particular historical interpretation. His preservation work can be valuable while individual conclusions remain open to examination. That is the standard I want applied across the board: give everyone what the evidence supports, take nothing legitimate away, and add nothing merely because someone repeats it often enough. The irony is that attempts to enlarge one legacy can eventually accomplish the opposite. Every public correction gives people another reason to approach the next recollection cautiously. The increasingly expansive title gives people another reason to return to contemporary documents. Every historical corrective that becomes too absolute invites another generation of historians to correct the corrective. None of this is necessary. The accomplishments are already present and accounted for.

Maybe there is a lesson here that extends beyond Petersen, Kuntz, Morgan, Gygax, Arneson, and Chainmail. The pioneers of tabletop and computer gaming were human beings. They weren’t prophets, saints, or perfectly reliable narrators. Neither are the historians, documentarians, bloggers, YouTubers, or writers attempting to reconstruct their lives today. That’s why the history should never belong completely to the last person alive who remembers it, nor entirely to the historian who arrives afterward and assembles the surviving pieces. Listen to the witnesses. Treasure their stories and record them while we still can, but compare those stories against contemporary documents, publication credits, manuscripts, correspondence, dates, and the recollections of other participants. When those things conflict, acknowledge the conflict rather than forcing the evidence into whichever creation story we happen to prefer.

Perhaps Chainmail is the oracle after all. Not because everything came from Chainmail, but because every attempt to simplify the creation of Dungeons & Dragons eventually seems to lead us back to that little medieval miniatures game and the tangled network of people, ideas, rules surrounding it.

There was no single road to Dungeons & Dragons. There were a dozen little roads crossing between Minneapolis and Lake Geneva, between wargames and fantasy, between Braunstein and Blackmoor, between Chainmail and Greyhawk, until eventually they met. And in 1974, somebody finally put three little brown books in a box.

Some legacies don’t need improving. They only need preserving.

Additional Sources