What is considered a beginner rating?

a 2000 elo player will beat easily 98% of all the players here.

a truly advance player is a beginner against the best software that cost 100$

and all those software are beginner compare to alpha zero....

and alpha zero is a beginner against the next best software of 2040 running on a quantum computer with 1000 qubits.....etc etc etc........

Colby is right. His claim remains unrefuted.

Improve in the rating ladder:

His point is irrelevant because 98% of the players here are crushed by anyone with a 1800+ rating.

And 99.9% of those players will never even play in REALITY against a true expert!

It's all relative like he says, but it's impressive when you see a GM preform well.

I played in a simul against a 2600 GM. There were about 50 people, and I knew that about 10 of us were 1900-2100 strength.

He won every single game, plus if he beat you quickly, he allowed you to set up the board and keep playing. I played two games, so did the 1900 player beside me. We all lost.

AND he even walked into a nasty line I'd prepared at home. When the tactics erupted he paused at my board for about 10 seconds, and unsurprisingly didn't find the engine move, but it and his subsequent moves were good enough to beat me.

Anyone who's played a simul knows it's really difficult seeing position after position. Your analysis becomes superficial. Even with "superficial" play he didn't even give up 1 draw.

Yeah it's all relative, Carlsen could give a small simul against 2600s, but when you see it first hand it leaves an impression on you, that you really are bad at this game

When I’m on a losing skid, I approach 1800. I am a beginner.

An IM completely destroy a 2000 elo player very easily....it;s like a Formula car against a guy on a bike.

No need to be a GM to crush a 2100 player.....

Eh, 2000 vs 2400 isn't a 100% thing. 400 point upsets sometimes happen.

But sure, it's very rare and overall the IM's skill and knowledge is tremendously higher.

I’ve given up a 500 point upset three times OTB to the same guy (and to two others).

I had a dead lost position against an 800 rated player out of the opening.

And no, he wasn't underrated. He lost all his games that tournament.

And no, I wasn't underestimating him, at least not that I know of. I went for some early tactics and won material... but the compensation I didn't foresee was enormous.

Luckily he was 800 for a reason and I eventually won.

Crap. I forgot about that 500 point upset when I was only 1400 and lost to a 900. Went 0-5 in that 1998 event. In 2012, I finished second in the same event—behind IM John Donaldson, and in front of an NM and several experts, including the one I beat in the last round.

Well, you dont need to wait quantum computer to beat Alpha Zero.

Current T60 Leela running on 4x 2080 Ti GPU in chess.com computership is approx +50 to +70 elo above Alpha Zero running in 4x TPU in 2017. 

How do we know?

1. New Leela Neural Networks are much bigger than A0 and much better AI training methods than they used in 2017.

2. Current Leela perform much much better than what A0 did against Stockfish 8 in 2017.

DeepMind released astounding results from an updated version of the machine-learning chess project today.

The results leave no question, once again, that AlphaZero plays some of the strongest chess in the world.

The updated AlphaZero crushed Stockfish 8 in a new 1,000-game match, scoring +155 -6 =839.

155-6-839 is a 57% win percentage.

CCRL 40/15 gives me a rating for Stockfish 8 = 3379

ELO tables link a 57% to a 47 to 53 rating difference.

So at best this performance is roughtly 3432

Putting DeepMind 60 points below the best stockfish on the CCRL 40/15 list.

155-6-839 is a 57% win percentage.

CCRL 40/15 gives me a rating for Stockfish 8 = 3379

ELO tables link a 57% to a 47 to 53 rating difference.

So at best this performance is roughtly 3432

Putting DeepMind 60 points below the best stockfish on the CCRL 40/15 list.

Of course this is just a (reliable) estimate, I'm mostly trying to point out that people go a little overboard with all the NN worshiping.

AlphaZero also bested Stockfish in a series of time-odds matches, soundly beating the traditional engine even at time odds of 10 to one.

In additional matches, the new AlphaZero beat the "latest development version" of Stockfish, with virtually identical results as the match vs Stockfish 8, according to DeepMind.

Today's release of the full journal article specifies that the match was against the latest development version of Stockfish as of Jan. 13, 2018, which was Stockfish 9.]

The machine-learning engine also won all matches against "a variant of Stockfish that uses a strong opening book," according to DeepMind. Adding the opening book did seem to help Stockfish, which finally won a substantial number of games when AlphaZero was Black—but not enough to win the match.

https://www.chess.com/news/view/updated-alphazero-crushes-stockfish-in-new-1-000-game-match

Pfft, when (the outdated version of) SF had an opening book DeepMind only scored 54% which is an Elo of ~30 points higher.

Neural networks are amazing, but the strength of their chess playing is completely overblown.

Now give SF all the bells and whistles. Give it powerful hardware with an opening book plus access to endgame tablebases. Of course they don't do that because a they might not even win, but if they did, it would be something like 51% to 49% which wouldn't make good headlines.

Give the same opening to A0 and SF is dead.

Do you know how many years it take to get to 2500 ? a beginner cannot be someone that worked hard daily for more then 10 years.