"OO" algs

[Labano]

Are "oo" algs worth it? I actually like them but many people don't recommend it.

 

[AlgoCuber]

It's just preference. If you like them and it's faster for you, use them. Otherwise, don't.
Probably the main reason why OO algs are not universal is that it may be uncomfortable for some.

OO algs are just algs that are theoretically the fastest. They're actually "subjectively optimal" rather than "objectively optimal," IMHO.

 

[Labano]

I can do this Gc perm faster than standard alg: The BEST Gc Perm Algorithm! GAN Predicted The Future

 

[abunickabhi]

Are "oo" algs worth it? I actually like them but many people don't recommend it.

OO algs are not worth it actually.I have tried them out and there is a large chance of fumbling it.

 

[Labano]

but some algs are worth it like for p shape olls

 

[DuckubingCuber347]

OO algorithm's are generally not worth it due to the high risk factor but the easier ones like wide-Gc and RUfS F-perm are extremely fast and the risk factor can be greatly reduced with practice. It's really just depends on your turning. Not everyone can use OO algs and some can.

but some algs are worth it like for p shape olls

What is the OO P-shape you're referring to?

 

[Labano]

The S slice alg

OO algorithm's are generally not worth it due to the high risk factor but the easier ones like wide-Gc and RUfS F-perm are extremely fast and the risk factor can be greatly reduced with practice. It's really just depends on your turning. Not everyone can use OO algs and some can.


What is the OO P-shape you're referring to?

 

[DuckubingCuber347]

The S slice alg

That's the standard alg though. If anything the OO alg is, S' r' F' r U f U F'. Just because there's an S move that doesn't make it OO.

 

[Labano]

before people used RUF isn't it?

btw from where do they get these algs?

 

[AlgoCuber]

btw from where do they get these algs?

Alg generators like Cube Explorer.
Also, treat this as a forum and not discord.

 

[ruffleduck]

"OO" is a meme term. Such a concept is silly.

For each individual, some algs are faster and some are more consistent. Seek a balance.

To answer your question on how algs are genned, it is usually done by playing around with the cube or using a program like Cube Explorer.

 

[Labano]

Thanks, also sorry I didn't know that

 

[OreKehStrah]

The problem with labeling algs as OO is not that they aren't OO but because they are only OO in a certain metric, such as movecount. It would be more accurate to call them" OO* , *optimal by xyz metric" but that's not catchy

 

[Labano]

So OO means High risk High reward.

 

[qwr]

There is an analogy to statistical models where one needs to specify carefully which metric a model uses. For example, for unbalanced classification tasks, accuracy is almost certainly not the right metric for the task at hand, and AUC has problems as well. Nonetheless, we can consider Pareto optimal algs in terms of several metrics (say, movecount, execution speed, risk) and get rid of all Pareto dominated algs.

 

[Labano]

I actually didn't understand

 

[BenChristman1]

In my personal opinion, some “OO” algs are worth it, although they aren’t actually “objectively optimal.” Algs have gotten so developed that for many cases, there are multiple good choices for which alg to use, it is just all about what you prefer.

OO algs are not worth it actually.I have tried them out and there is a large chance of fumbling it.

Doesn’t 5-style use a lot of weird moves and fingertricks? It’s no different, just that the moves that were previously thought of as impossible (such as S and E moves) are now being applied to speedsolving as well.

 

[GenTheSnail]

I think the best way to determine an alg isn't the peak speed that you can reach on it, but the peak speed you can consistently and reliably reach on it. It doesn't matter if you can do an RUS uperm in 0.44 if you mess up and do 1.4 every other solve. In that case, an MU or RU that you can consistently do in 0.6 is going to be a better choice.

Try executing an alg for an ao25 a mo25 (without deleting any mistakes) and see which one is faster then.

 

[AlgoCuber]

I think the best way to determine an alg isn't the peak speed that you can reach on it, but the peak speed you can consistently and reliably reach on it

In my personal opinion, some “OO” algs are worth it, although they aren’t actually “objectively optimal.”

Completely agree. For example, OO Gc perm (R2' u' R U' R U R' u R2 f R' f') is the fastest for me to execute. However, I find myself doing way more consistently with the regular Gc perm (which I execute like this). However, on Gb perm, the OO version is pretty much always faster for me. I also hate algs with S slices because I can never seem to do it consistently. I know many on Discord and on this forum that will completely disagree with this, and that just explains why OO algs are not objective.

As I said before, OO algs are more like "theoretically optimal" than "objectively optimal." If everything goes perfectly, an OO alg is probably faster. However, most of the time that is not the case.

 

[xyzzy]

There are multiple wildly incompatible definitions of "OO" floating around.

1. Algs that use a lot of moves that used to be considered bad (e.g. B moves, S moves).
2. Algs that are risky but potentially fastest.
3. Algs that are optimal by some objective measure.
4. [insert meme cuber here]'s favourite algs.

(Probably not exhaustive. Wouldn't surprise me if someone else wanted to chime in with their own meaning of "OO".)

None of these is relevant to you on a personal level.

You should pick algs that are the best for you. If you don't know which ones to pick (especially if you're a newbie and don't have a wide fingertrick repertoire yet), then you can look to more experienced cubers' choices. But you still shouldn't treat them as gospel, and you should be open to switching to other algs.

Alg speed consistency is also a very subjective decision to make. If you're aiming to break records, you absolutely should go for high risk high reward choices. If you're looking to podium at a local comp, your strategy should depend on how fast the other competitors are. If you're trying to improve your average, you should go for algs that are the fastest on average, not necessarily the ones that have a lot of peak potential but are extremely risky.

---

Try executing an alg for an ao25 (without deleting any mistakes) and see which one is faster then.

If you're measuring alg execution times, you should almost always just use the mean rather than cutting out best/worst. The outliers matter, because the alg isn't the full solve; a 5th-percentile-bad execution time on, say, an RUS U perm alg doesn't mean that the solve as a whole is also going to be 5th-percentile-bad and be excluded from the average.

---

Also, treat this as a forum and not discord.

I mean, I have a tendency to wall-of-text here, but short posts should also be welcome, methinks.