Justice Toolbox Rating System

Justice Toolbox rates attorneys on a five-star scale based on their past performance in court cases. We have collected millions of court records and analyzed them with our artificial intelligence program to determine which cases an attorney has won or lost. We then rate attorneys based on two factors: (1) estimated win rate and (2) experience.

We chose to use estimated win rate, measured as the percentage of cases won, as a rating factor because a good attorney should be able to produce successful results. It is not possible for an attorney to win every case, but, statistically, over a long period of time a winning pattern should emerge for good attorneys. Win rate is calculated as wins divided by the sum of wins and losses, without regard to other outcomes like settlements. The win rates computed are estimates because they are calculated by an artificial intelligence based on court records.

We chose to use experience, measured as the total number of cases an attorney has taken of a particular type, as the other rating factor because a high win rate over a few cases is much less meaningful than a high win rate sustained over a large number of cases. A large number of cases indicates that the attorney's win rate is a real pattern and not a fluke. In addition, an attorney who has taken more cases will naturally be more skilled due to repeated practice, and, therefore, experience remains an important factor even when considered independently of win rate.

The Ratings


The attorney is outstanding in both win rate and experience. This attorney's estimated win rate and experience are both in our highest percentile grade when compared to other attorneys for the same types of cases and taking into account whether the attorney primarily represents consumers or institutions and businesses. A 5 star rating is difficult to achieve and some practice areas may have no attorneys with a 5 star rating.


The attorney is excellent. Either the attorney's estimated win rate or experience is in the highest percentile grade and the other is in our very good percentile grade.


The attorney is very very good. Both the attorney's estimated win rate and experience are in our very good percentile grade, or one is in our highest percentile grade and the other is in our good percentile grade.

3.5 or

The attorney is good. Both the attorney's estimated win rate and experience are at least in our good percentile grade, or at least one is in our very good percentile grade or highest percentile grade.


Attorneys may be unrated when we do not have enough data to compute a rating. No rating does not mean the attorney is bad. Often attorneys are unrated in areas where we have identified few recorded cases handled by them. We recommend asking the attorney for more information about their experience in this area. They may be able to provide information about additional cases or outcomes they've achieved.


Both estimated win rate and experience are measured by comparison with other attorneys. We calculate attorneys' estimated win rate and experience percentiles, which measure the percentage of attorneys that this attorney has a higher value than. Percentiles are computed separately for each case type, and also separately based on whether the attorney primarily represents consumers or institutions and businesses in the case type. This is important because win rates can vary widely by case type, so attorney win rates should be compared within the same case type. Likewise, some practice areas involve much higher win rates for either the consumer-side or institution/business-side in the case, so it is also important to compare attorney win rates with others on the same "side."

Prior to computing the percentiles, we remove outliers in order to avoid results being thrown off by large numbers of attorneys who have a 0% win rate with 0 wins and 0 losses or a 100% win rate with a small number of wins. This involves removing the attorneys who have 0 wins and 0 losses and the attorneys who have a small number of wins and losses (25th percentile and below) prior to computing the win rate percentiles. We also treat attorneys with a 94.5% estimated win rate or higher over a large number of cases equivalently to attorneys with a 100% estimated win rate over fewer cases to keep things fair for attorneys who may have a few fluke losses.

Ratings are awarded per case type. An attorney who is skilled in one area is not necessarily skilled in other, unrelated areas. Different case types require different knowledge, skills, and expertise. Therefore, we rate attorneys based on estimated win rate and experience per case type, and these ratings are completely independent of other case types.

However, there are three exceptions to the per-case type ratings: "Overall," "Civil," and "Criminal." Ratings given for these areas are the maximum rating of any of their subfields. This makes sense because an attorney can still be excellent "Overall" if he or she practices almost entirely in one area of specialty. Many attorneys specialize. It would not make sense to give an attorney a lower "Overall" rating based on areas where they rarely take cases or have little interest. "Overall" includes all cases. "Civil" includes all cases within the civil system, which generally covers all cases where two individuals or businesses are in a lawsuit for money or other relief. "Criminal" includes all cases in the criminal system, which is when the government prosecutes an individual for a crime and a fine or jail time is at stake.

All ratings are computed using only the data currently available to us. We do not have access to all court records in the United States. While we have millions of records, these are only a fraction of all the court proceedings in the country. We may be missing wins and/or cases for some attorneys if they occur in geographic areas where we have no data. Moreover, all outcome data is computed automatically by an artificial intelligence program without verification by a human.