Sam Anschell

277 karmaJoined Jan 2022


Thank you for this thoughtful comment, Tyler - I appreciate your perspective and I think it will help readers improve their decision-making. 

On point 1: I suspect this is less true for entry level roles, especially those that don’t specify an advanced degree or technical skill requirement. But it’s valuable to know that this was your experience when reviewing applications, and this updates my opinion.

On point 2: I agree there are more early career EAs looking for EA jobs than entry level EA job openings at any given time, but I disagree with the conclusion that early career EAs should apply to fewer EA jobs.

  • It seems hard to get the same level of mentorship, relevant skills, at most “non-EA” orgs. If you’re working as e.g. an SWE for FAANG, your employer’s incentives are to invest in your professional development insofar as that increases your productivity and job satisfaction (for retention). If you’re working as e.g. a researcher for the Center for Global Development, your employer’s incentives are to invest in your professional development to maximize your lifetime impact on your & CGD’s shared mission (agnostic as to whether you achieve that impact while working at CGD or elsewhere). 
  • You might be a great culture fit, or have a particularly relevant background that headhunters or EA orgs wouldn’t know about unless you actively apply.
  • Of the current entry level roles on 80k’s job board, there are a pretty diverse array of functions. I agree that applying to all 224 wouldn’t be a good use of time, but regularly checking back in and applying to promising leads seems like a good idea to me (especially keeping an eye out for jobs matching any specialized background/knowledge base you might have, like in policy, academia, development economics, ML, infectious disease, etc.)
    I personally made the mistake of applying to a bunch of stuff all at once, feeling disappointed about not getting anything, and giving up until I felt motivated to apply again later. I think I should have applied consistently and internalized that the EV of an application working out (in my opinion) makes the application process worth it. 

    The post and this comment each have the tacit asterisk that individuals' situations are unique, and that I'd be excited for the hypothetical average early-career reader to apply for more EA jobs/get more involved with EA projects on the margin. Some people don't have the time/financial runway to apply for lots of jobs or volunteer for EA projects, and I hope my advice isn't perceived as one-size-fits-all.

On point 3: Couldn’t agree more :) 

I know 80k used to recommend careers in management consulting pretty strongly for skill development, and these days my read is that they’re starting to recommend more “direct work” roles right away. I’d love to learn 80k's current view, and I would recommend that a reader weight the perspectives of the professionals (those with recruiting/hiring experience like Tyler, and those who research the EA job landscape like 80k) much stronger than my perspective when deciding what makes most sense for their personal career circumstances. 

I love the spirit of this post and I agree with almost everything Charles said. My only difference of opinion is that I don't think people are taxed on gift income, so long as a giver provides under $16,000 in a year to any single individual and under $12.06MM in their lifetime. Letting participants keep the tax savings could be a polite way to compensate them for their involvement without dampening their altruistic glow.

I'd be excited by a project that explored surrogate donation from employees of these companies to multiply the impact of our giving. I expect the cap on Meta's Giving Tuesday match will barely scratch the surface of EAs' personal donations this year. My preliminary thoughts to scale this idea would be:

  • Learn  the fine print of the programs. Some employers explicitly disallow donating money that is given to an employee for this purpose.
  • Try to form a nuanced understanding of the long term incidence of doing this. I imagine many companies would discontinue or scale back their charity benefits if it was exploited in this way. 
    I suspect that it would still be net good because of which charities would get bonus funding, but companies price in the fact that most employees don't utilize donation matching & I suspect they would adjust their offering if it were used more.
  • Focus on employees of companies with the most generous matches first. Possibly establish a fund where EAs could pool their money and funnel them through each surrogate donor to max out the donor's gift cap.

I want to reiterate what Charles said about disassociating this from EA, and I recommend speaking to a(t least one) lawyer before making any moves. This feels like the type of ends-justifying-means action that is righteous in intent, but underhanded in public appearance.  

Feel free to reply here or PM me if you'd like to flesh this out. I think it's a great idea, thanks for sharing!

Thank you for the thoughtful writeup! I knew very little about PhD programs before coming across this post - now I feel like I have the lay of the land :) 

It seems like the GRE isn't a dealbreaker for getting a PhD, but for anyone interested in 80/20'ing their GRE prep I'll shamelessly bump my GRE advice post (thanks to Robi Rahman for his help)!

To me it seems likely that the majority of good opportunities will still be around in June. If there is something like $7,500 promo EV up for grabs in NJ now, I would estimate with 70% certainty there will be more than $4,000 promo EV by June. 

One wrinkle I want to make sure you're aware of is that creating, funding, betting with, and withdrawing from ~ 17 accounts will take a significant amount of time and needs to happen over the span of at least 4 days (for your promo credit to post, and for any withdrawal follow-ups to register. I recommend staying for 5+ days and starting this process the day you arrive.) 

My best guess is that having never done this, you'll need to spend 15 hours to complete every worthwhile promo. If you've got a full schedule on your NYC trip (especially including time commuting to/from NJ), it may be best to schedule a separate trip. If you have some time but not the full 15 hours, maybe the best course of action is to do only the best offers and accept higher variance with a smaller sample size. 

Thank you Vivian, you're absolutely right. I'll update the "qualifiers to participate" section to make this much clearer, especially for participants living in states with a small number of legalized online sports books. 

Strange, I was able to withdraw to a debit card or through PlayPlus (which I can then transfer to a bank account). Your withdrawal options may vary depending on what method you use to deposit. 

I've discussed this idea with a few people. It's unclear whether it is legal to send someone money for making online sports bets per the Wire Act. To my knowledge, if this is illegal no one has been convicted (similar to how playing online poker in certain states is grey-area illegal but the law has never been enforced). 

I "may have" loaned two friends $15,000 to participate, but I'm not sure whether I'd be comfortable setting up an EA fund for this:

  1. Even if there is a way to learn that this is completely legal, the risk to Effective Atruism's image if just one participant became addicted to gambling through this funding may eliminate all expected gains.
  2. There is a ton of trust required, and involving a loaner/contract lawyer's time decreases $/hour of this project. 
  3. Someone strapped for cash seems less likely to itemize on their taxes (though I could certainly be wrong) and thus would make less from this. 

I like your thought to scale this; EAs can raise more money by funding people they know and trust, so long as they do it discreetly and understand that there is some chance they are breaking the law. 

I also think that by the standards of the best opportunities for direct work, $500/hour worth of impact is somewhat low. Per a call with Alex Holness-Tofts, community building at college campuses is conservatively estimated to create $80,000,000/year in impact, and phone banking during the 2020 election may have been worth as much as $9,000/hour. Right now, campaigning for Carrick Flynn likely does more good per hour than these sports bets. So I'm not sure how valuable it would be to have an EA institution push cash-strapped EAs to do this rather than something else.

When I bet with SI it was similar to DraftKings, but now they’ve changed the promo to be a bit worse for those taking the standard deduction. You have to place 30 $10 bets to unlock 30 $10 free bets that do not return they’re stake

Strategy should be short odds for the cash bets, long odds for the free bets. Still worth doing, just a bit cumbersome. Thanks for spotting that!

Napkin math says no if you’re taking standard deduction but yes if you’re itemizing. Promo EV is about $1600, and you have to place $6,000 worth of bets to unlock it, so if your bets outperform a ~27% loss rate, this is profitable.

Be sure that the cash bets you place are no shorter than -200 odds because -200 is the minimum odds to qualify towards to rollover requirement. Also, the bet you place with the site credit should be on long odds because the free bet does not return its stake.

These opportunities exist because each book sets its odds independently. Sports books respond to new information asynchronously and the books' different models can overweight the likelihood of certain outcomes and underweight the likelihood of others. From my experience, BetMGM and Fanduel often give favorable odds on favorites (short odds), and DraftKings, FoxBet, Barstool, and Twinspires give favorable odds on longshots. 

Another reason these arbitrage and +EV betting opportunities exist is that US sports books are new, and the operators are prioritizing growth/debugging/marketing above the accuracy of their betting lines. Offshore sports books' odds are far more similar to one another than US sports books'.  I would predict that by 2025, the opportunity to make >5% +EV bets on US sports books will be a shadow of what it currently is.

These opportunities aren't being taken advantage of for the same reason that blackjack players aren't all card counters: It's stressful to bet large amounts of money on high-variance events even when you know you're a mathematical favorite. And once a casino is sure you're a winning player, you're no longer welcome to play. 

For anyone with the stomach and the bankroll for making hundreds of these slightly profitable bets per week, the opportunity is incredible until you get limited. With 60% confidence, I would estimate that someone with a liquid bankroll of $100,000 could profit between $40,000 and $200,000 within four months of advantage sports betting (8-12 hours/week). All they would need to do is bet on all >2% +EV betting opportunities that OddsJam suggests (with a tight market width/confidence interval), and use OddsJam's Kelly Criterion calculator to responsibly size their bets. 

I would recommend that anyone considering this path be selective about suspicious-looking bets. Player props and alternate lines can be red flags for books to limit you quicker. If I lived in a state with legalized online sports betting I would 100% be pursuing this; I'm happy to answer more questions or set up a Zoom meeting via Calendly if anyone is interested!

Load more