This post is an oblique response to http://adeptplay.com/actual-play/dd-habits-and-culture and http://adeptplay.com/seminar-hearts-minds/conversation-dd-play-culture . I am not writing to give a definition of OSR, but rather describe my experiences over a number of years and how I play in and run games that are often classified as OSR, but also various other editions of D&D.
I am talking about a specific creative agenda or style of play or purpose to play or whatever terminology one prefers. It could be described as problem solving play (or challenge-focused play) with a focus on fictional challenges, exploration, strategy, tactics and logistics; or wargaming on a commando squad level. For me, the fun of playing in this way comes from exploring the unknown and creative problem solving, as well as similar fun one gets out of various board and computer games (which are much more limited in scope by necessity of not having a human referee adjudicating whatever actions one tries). I've been running games in this style using various homebrew rules systems (some of which would be obviously classified as OSR, some might not), Pathfinder 1 and D&D 5. I've played it using Lamentations, Old school essentials, D&D (Finnish red box and Finnish B/X), a homebrewed system, a homebrewed descendant of BECMI and Praedor. And maybe Basic fantasy roleplay or some other retro-clone.
The games are I've played in have almost always been sandboxes. A couple of the convention games have been one-shots with no connection to anything, but most con games have had a persistent world and possibilities to play in the same one for several sessions. What sandbox play means here is that there are several different things going on in the game world, including several obvious adventuring possibilities. You can engage or not with any, but in longer-term play it might have consequences. You can scout out locations, choose to go in or not, scout another one, return and leave, and so on. There is also maneuvring outside the locations themselves; enemies might escape from a location if players go in with force and then leave, or they or the characters might set up ambushes or track down their enemies. Some locations, like ancient tombs, are essentially static, while others, like lairs or castles of intelligent beings, contain active actors.
The default gameplay loop is what one would expect: go into a dungeon, try to survive and find treasure (or reach some other objective), and then get out alive. As players get their characters more established in the game world, they typically start having different kinds of goals - resettling an abandoned town, starting a silver mine, and so on. Sometimes the world imposes things on them - an army approaching, someone messing with a holy treasure and a god imposing a quest on them. Sometimes characters are retired as they find a non-adventurous place for themselves in the world.
The key principle for this kind of play is that the game master has to be as neutral as possible with respect to the survival and success of the player characters. Apocalypse world had the game mastering disclaiming decisions making (about certain dramatic issues), whereas here the referee is disclaiming decision making about the fate of the player characters. One way to do this would be to use a rules engine that defines many things carefully (this is what West marches of Ben Robbins did), but I prefer a different take. It is essentially a combination of several techniques with a focus on establishing trust on this neutral refereeing. Rolling dice in the open and saying what they are rolled for is a basic technique; for example, pushing the giant sloth down the shaft would be difficulty so and so strength roll, given there are two of you and supposing it does not notice you while you charge it; or you would expect 1/2 chance of finding a carefully hidden ("typical") secret door in the room if all of you search for twenty minutes or if one is on guard but the rest search for about twentyfive minutes.
The way I make these judgments about difficulty levels relies on several things. First is going by the fiction and how the world works, to the best of the group's understanding. Being explicit about the difficulties allows the entire group to voice whether the numbers are reasonable or not and whether there are misunderstandings and whether my estimates are good. Since I play with people who know more than me about various matters, we can benefit from this shared knowledge pool, learn, and make the game world more vivid. As I play more with a given group, things get more codified and there is less need to be explicit. The second procedure is reading what the adventure or preparation says; the third is checking if the rules system has a way of resolving the matter.
More generally, to referee in a neutral way, I always think what might happen in the world, and in case of uncertainty, consider probabilities for different options and roll a die. I use prepared adventures (mine or by others) as they specify threats and treasures; I roll random encounters to decide whether something is met or not; I make reaction rolls to determine whether someone is well or ill disposed; I roll morale check to determine whether someone fights on or retreats. Sometimes the rolls are spontaneous; I might roll d3 to determine whether some group fortifies, runs away or makes a counter-attack after the player characters caused heavy losses to them and gave them time to react, unless the political dynamics or the personality of their leader determine otherwise.
In one of the linked videos there was talk about hesitation and confusion before opening a door. This is something I would expect from an inexperienced group; the game master (nor any of the players) is not making clear the difference between planning and acting and none of the players takes leadership. The first unclarity I have experienced now and then, not really with OSR as I have not played that with an inexperienced group, but most recently with Burning wheel. Very frustrating. Since I see the game as a wargame, teamwork and leadership are interesting and relevant parts of play, so the second is a case of inexperience as a player in this style of play or a skill one has not mastered yet. Much like in other (real life) team work contexts, if the team is undecisive, one has to guide it into a more fruitful mode of action. As a game master, I would say, after a while, that I interpret it so that the characters are having a discussion about what is happening (allow the players to disagree or agree) and this has taken, say, ten minutes or so (with maybe a reminder about how torches or random encounters work if appropriate); do you continue or do something else? Greater player skill (in teamwork skills in this case) leads to play that is more fun and engaging, and those players also get better results in game due to more functional decision making.
I would say that my role as a game master is to act as coach to new a group (an experienced group learns by experience, not coaching) and as a neutral referee in any case. I am not against the players; I am hoping they do fine against the game world and adventures. I have made the world and chosen the adventures, but not for this group of players or characters; I have added a hopefully healthy mix of difficult and easy, cruel and forgiving adventures, by different authors and from different game systems, and over a long period of time. Some places have been explored (and randomly restocked), some not. I do not typically know the level of the characters that are active at the moment, except vaguely (this one is level one or two, that is around six). This all makes sure that I am not tailoring any content difficulty-wise.
The game masters I play with might or might not play in completely the same way, but the games have been sufficiently compatible that I have not been surprised by how they run their games. In one game we went to an adventure location that had been emptied, figured out what had happened (from the place and notes left by the previous group; this was a con game with a persistent world with several sessions, persistent characters but obviously shifting players). Little loot, little action, fun detective work. Next session with a different group we used the information, figured out a shopkeeper had all of the loot in their store, did some social engineering and got the town guard to surround the place, getting a good share of the treasure with little risk. My character retired after that, being the old man they were. In another session in the same continuum we went to a different adventure location, again mostly looted, wisely left a tomb undisturbed, and found a hidden treasury (there was a statue with three heads, two of which looked at doors; the secret door was where one assumed). I have also lost characters due to recklessness (climb down a cliff without a torch and then there was a guard down there; be eaten by a magical owlbear when I was not good enough at hiding; these in completely different games), but I have not, even once, felt that it had been unfair; I made my choices, stupid or not, the dice added friction or helped along, and then there were consequences.
On a more tactical side, in a different game, my pseudo-mongol mounted archer (1st level) faced two orcs and fifteen goblins and managed to drive them away almost single-handedly while killing an orc and a goblin due to the mobility given by the horse and having a bow (as did the orcs). The rest of the party was fortifying the manor we were here to explore. There was definite risk-taking; I was hit once and it was good luck to slay the orc in heavy armour, acting immediately while they were still confused about my appearance, and changing my angle of attack and not staying put.
I have not discussed the role of the written rules system that much. Much as in other traditional games, the style of play is more important than the written rules system. In this context, the rules system gives the basic mechanical framework and basic chances of success of various actions, as well the power level and superpowers of the characters. The most important rules change I use is experience from achieving player-chosen goals (with 1 experience point per gold piece of treasure recovered being the default unless you have better things to do; gets the game going nicely); since this is a game of player skill, the players need to set their goals and experience is how we measure good play, so the experience rewards should be something the players can predict and reason about, to some degree.
There are many other things one could discuss; how some people simply do not believe that this is in fact the way we are playing (leading to new players facing more frequent character deaths unless coached by more experienced ones), how one can see a marked increase in skill as people play, how many skills transfer from game master to another, how the game works fine with one or two active players taking leadership and others mostly just hanging on and even better when everyone is active and engaged. There are also all the fascinating questions about how to best play the game, like matters of scouting and being on watch and how very careful and meticulous play sometimes has a great cost due to attrition of resources or due to intelligent enemies taking initiative.
This is a very general text; please let me know if there is anything particularly interesting here, or missing from here.