I am not particularly fond of playing one game session and going up a level. That hardly qualifies as "earned," to my way of thinking. (E. Gary Gygax as Col_Pladoh on Dragonsfoot forum)
The above quote was posted 30+ years after the original rules were printed. It supports "designer intent" in this answer.
For a point closer in time to the original rules being published, an article by Gary Gygax in Strategic Review (April 1976, p. 22-23, D&D IS ONLY AS GOOD AS THE DM) addresses earning levels. A theme found throughout this article is:
It isn't better to be lucky than good.
Successful play of D & D is a blend of desire, skill and luck. Desire is often initiated by actually participating in a game. It is absolutely a reflection of the referee’s ability to maintain an interesting and challenging game. Skill is a blend of knowledge of the rules and game background as applied to the particular game circumstances favored by the referee. Memory or recall is often a skill function. Luck is the least important of the three, but it is a factor in successful play nonetheless.
The article emphasizes that player skill is important. Gygax refers to unearned levels, and blames that on Dungeon Masters who are not doing a good job. He further states that D&D is enjoyable if the players have to earn their levels. Earning levels should be a result of player skill above other considerations.
When read from top to bottom, the intent is clear: players should earn each level up, not be "given" more reward than appropriate for that level.
By requiring players to work for experience, to earn their treasure, means that the opportunity to retain interest will remain. It will also mean that the rules will fit the existing situation, a dragon, balrog, or whatever will be a fearsome challenge rather than a pushover.
The article mostly address high level play, rather than lower levels where an XP spike is more likely to present a chance to earn enough XP to advance more than one level during one adventure/session. The consistent theme is that the rate of level gain should not be rapid, and that it should be a challenging process.
Experience points are heaped upon the undeserving heads of players, levels accumulate like dead leaves in autumn, and if players with standings in the 20’s, 30’s and 40’s of levels do not become bored, they typically become filled with an entirely false sense of accomplishment, they are puffed up with hubris. As they have not really earned their standings, and their actual ability has no reflection on their campaign level, they are easily deflated (killed) in a game which demands competence in proportionate measure to players’ levels.
- Note: this "rate of level gain" theme was again spelled out in the Companion Rules (BECMI):
"After reaching name level, characters should gain a new level of
experience or every 3 to 8 successful adventures." ( Dungeon Master's Companion p. 2)
More E.G.G. on how a DM should make a challenging rather than a giveaway campaign:
When players no longer have reams of goodies at their fingertips they must use their abilities instead, and as you will have made your dungeons and wildernesses far more difficult and demanding, it will require considerable skill, imagination, and intellectual exercise to actually gain from the course of an adventure.
Player skill is the issue, not "character attributes" nor "character luck."
When it is difficult to survive, a long process to gain levels, when there are many desired items of magical nature to seek for, then a campaign is interesting and challenging. Think about how much fun it is to have something handed to you on a silver platter — nice once in a while but unappreciated when it becomes common occurrence. This analogy applies to experience and treasure in the D & D campaign.
There is a "metric" presented on expected advancement:
It is reasonable to calculate that if a fair player takes part in 50 to 75 games in the course of a year he should acquire sufficient experience points to make him about 9th to 11th level, assuming that he manages to survive all that play.
Keeping in mind the "1 GP = 1 XP" model (plus points for monsters from the XP table) and how the random treasure tables in Monsters and Treasure (and in Greyhawk) provided spikes in the GP value of treasures found, this article shows that the designer's intent is that level progression should be a gradual process.
With that intent considered, the rule from Men and Magic fits as a means to that end: don't advance too quickly because that would not reward player skill at the appropriate rate. (At the philosophical level, "the journey's the thing, not the ending" applies as well if one considers what Rob Kuntz, a playtester from the beginning, has to say in his recent writings on FRPG's).
Fast forward three years to the publication of the AD&D 1e DMG. That same intent is reflected in Gary Gygax' detailed rules on advancement in level (AD&D 1e DMG p. 86).
The above analysis is not a direct statement of
"why we did that back in 1973/1974 when we first published the rules"
but it shows evidence of design intent by the publisher.
That general intent is reflected throughout the article on how hard it should be to earn levels -- even though much of this was directed at higher level advancement.
The acquisition of successively higher levels will be proportionate to enhanced power and the number of experience points necessary to attain them, so another year of play will by no means mean a doubling of levels but rather the addition of perhaps two or three levels. Using this gauge, it should take four or five years to see 20th level. As BLACKMOOR is the only campaign with a life of five years, and GREYHAWK with a life of four is the second longest running campaign, the most able adventurers should not yet have attained 20th level except in the two named campaigns. To my certain knowledge no player in either BLACKMOOR or GREYHAWK has risen above 14th level.
Can the designer's intent for putting that rule into print in Men and Magic be determined from EGG's post hoc analysis of the game's progress two years later, after the benefit of 4-5 years of play test by the designers and an ever growing legion of gamers and fans? I argue that the answer is yes1, particularly in light of his comment at Dragonsfoot.
Player Skill as a Core Value is a Carry Over From the Wargames D&D Grew From
The early form of D&D grew out of competitive table top miniatures wargaming where player skill was a core ingredient to winning a given battle. All of the fun, whimsy, adventure, magic and "spark" that was folded in from Arneson's innovations and collaboration with Gygax -- stuff we all take for granted now -- doesn't change how closely related to the wargaming model the original form of the game was. As stated on the box, and on the cover of Men and Magic, Dungeons and Dragons rules were "Rules for Fantastic Medieval Wargames Campaigns Playable with Paper and Pencil and Miniature Figures."
After a few years of playing in two different campaign worlds -- Greyhawk and Blackmoor -- the game went to print. (In fairly rough form). Enough play and play test had transpired to expose some of the game's quirks. From seeing that rule in Men and Magic, repeated in Holmes, repeated in BECMI, and later expressed more formally as the game grew, it is fair to conclude that they discovered during play--before the rules were published--that a player could get an XP spike sufficient to gain more than one level at a time, and that this "feature" was seen by one primary dev -- Gary Gygax -- as a "bug."2 The "bug" fix rule inserted on page 18 was preemptive, and like most of the rules in OD&D it was terse. The rule told you what, not why.
In the mind of at least one dev, E.G.G1, an XP spike does not reflect player skill.
It takes player skill to earn levels.
Level advancement must be earned.
Designer intent is later confirmed by the expanded treatment of leveling, and its limits, in 1e AD&D, and some years later in a Q&A.
1 For all of Arneson's influence on the game, when it came to getting the rules written down and published Gygax had a far larger share of impact, and in a lot of cases had the last word. (See Tim Kask's comments for how that played out as the game became more successful). The folks at Judge's Guild had similar frustrations to TK's with putting DA's Blackmoor material into publishable form. (Folks who played in DA's games were effusive in their praise for much fun they were, publishing issues being a separate category to fun).
2 My notional five dollar bet is that Arneson didn't feel strongly about this one way or the other. Can't back it up, however, so it's unknown.