If we absolutely MUST respond to rating inflation by making it harder to become a GM, the simplest solution is probably the best: Raise FIDE rating requirements by 100 points across the board at the beginning of 2020. If you're a titled player now, you keep your current title, no questions asked.
Simple
We could adopt a system similar to the Army regarding their Generals. That is to say you start off as a 1 Star General and try to work your way up to the 5 Star level.
Yeah, if only there were a system of titles to distinguish different levels of masters...
BTW: It should count for class players too: Expert = 2100, A = 1900, B = 1700 etc.
If ratings inflation devalues everyone's rating, then everyone should be effected equally by changes to the title system.
But the logical outcome of handling it that way is that by the year 2120 new players will start with a 2700 rating and you'll need 5000 rating for a title.
But the logical outcome of handling it that way is that by the year 2120 new players will start with a 2700 rating and you'll need 5000 rating for a title.
You're not shifting ratings, just titles, so the logical outcome is there would be fewer players in each category that was shifted... although I suppose shifting something like E class closer to the average would actually increase the members of that category 
As long as the numbers appropriately signify someone's level and the community is capable of interpreting them, it doesn't matter how big or small they are. A rating point isn't like a centimeter; it's not something whose value can be measured, standardized, and left alone. Periodically adjusting for inflation seems like a reasonable option.
I also think there's a tendency to talk about rating inflation and changing the rating system for top players without adjusting our (much lower) ratings for the supposed loss of value. If 2500 isn't a GM like it was in the old days, 1800 isn't what it used to be either. We're all riding in the same leaky ship, so let's ride in it.
But if players are actually better then there's no inflation. Players start younger, have access to more material, and play more often. Why wouldn't they be stronger?
If so, then the GM title is devalued in terms of rarity, not skill.
Btw I still like the idea of adding a Super GM rating as a simple way to improve things without creating any problems for those people who are now ranked as GMs ( ie as RGMs, Regular GMs ).
Hard to say after all the acquisitions. General Motors is home to Buick, Cadillac, GMC and Chevrolet
super duper gm
And will every Swiss and Israeli citizen automatically get a GM title when they turn 18 - whether or not they morally agree with the practice of playing chess?
I wholeheartedly agree with #8. I think GMs of today might be just as strong as GMs of the past.
Rating inflation and rating deflation (which arises more naturally btw) are two pitfalls to keep in check, but I don't think we fell for them.
In 2019, there are a lot more chess players than 50 or 60 years ago, and there are a lot more money in the system to sustain their study of the game, as well as more studying ressources (computers obviously, but also simply experience gathered as a community through these years).
The proportion of GMs among rated players is actually shrinking. It is also explained by casual club players tending to get a rating more often that they used to, but on the long run, it means a bigger community and inevitably leads to a larger elite.
If anything, I am for adding a new title rather than changing the existing ones.
If we wanted to base titles on rarity, as opposed to skills, the requirements should not have be based on rating in the first place, but on ranking.
If anything, I am for adding a new title rather than changing the existing ones.
If we wanted to base titles on rarity, as opposed to skills, the requirements should not have be based on rating in the first place, but on ranking.
Agreed to both.
The original Grandmaster, in 1950, was from the start meant to be a lifetime title, the initial 27 included old masters in their 80s.
How high should a ranking requirement be to give about 30 living great-grandmasters inclusive of the retirees?
For example, define a requirement as "Has been in top 5 of the world":
1969:
1970:
6) Geller (4)
7) Larsen (5-6)
1971: all old;
1972:
8) Polugaevsky (3-4)
9) Portisch (5-6)
1973:
10) Tal (2-3)
1974:
11) Karpov (2)
1975: all old
1976: all old
1977:
12) Mecking (4)
1978: all old
1979:
13) Timman (5-6)
1980: all old
1981:
14) Hübner (4-5)
1982:
15) Kasparov (4)
1983:
16) Ljubojevic (3)
17) Andersson (4)
1984: all old
1985:
18) Vaganian (4)
19) Beliavsky (5-6)
1986:
20) Yusupov (3-5)
1987:
21) Sokolov (3-4)
1988: all old
1989:
22) Short (3)
23) Speelman (4-5)
1990:
24) Ivanchuk (4)
25) Gurevich (5-6)
26) Salov (5-6)
1991:
27) Gelfand (3)
28) Bareev (5-6)
Hm. 22 years to collect 28 names. How many would have come in the 28 years since?
I dont understand. Who is complaining that there are too many grandmasters? Aren't there about 1600 in the world? That seems like a very rare accomplishment. Besides, isn't the current system of using ratings good enough to identify which grandmasters are better than others?
Bah, rating is overating... you list by elo then player like Caruana and Giri be stronger than Fischer yes?
A grandmaster should be the elite of the elite. The requirement to be a gm candidate should be to have an expected score of 2.5 or greater if you entered a round robin with the top ten players not counting yourself. During the same period as a candidate you should be required to produce 3 performance ratings in tournaments 10 games or longer of the world champions actual rating or higher. If you no longer meet the candidating requirement, your norms reset.
For Im's it should be the same except using the ten newest gm's as a benchmark for expected score and the third highest of the newest ten as a benchmark for necessary performance ratings.
In recent years there have been a number of complaints that there are just too many GMs these days and the title has therefore declined somewhat in prestige. We could adopt a system similar to the Army regarding their Generals. That is to say you start off as a 1 Star General and try to work your way up to the 5 Star level. So we could have a RGM level ( Regular GM ), then next a SGM ( Super GM ), also a FWC ( Former World Champion ) and of course also CWC ( Current World Champion ). Note to the FIDE: I expect a finder's fee for this terrific idea lol.