Methodology

My first try to identify Algodex's order book was to look for on-chain transactions on the algorand.application_call_transaction table and filter by app_id. I contacted the Algodex support team to obtain this identifiers which are:

• Algodex_ALGO_Escrow 354073718
• Algodex_ASA_Escrow 354073834

Results

From the daily tx_id count, given the data from February 9th is similar to data before (supposedly not launched) and the big increase in activity in February 10th, I will evaluate the order book on February 10th for the rest of the dashboard.

Conclusion

This was the most challenging bounty so far for me since there were lots of things I had to assume and really dig deep to find out. The first query on the algorand.application_call_transaction table was not really successful but that lead me to investigate further and with the help of Discord user @The Laughing Man I was able to obtain some results. I am really looking forward to check other users submission, specially the one granted with the grand prize, to see if my results are correct.

Introduction

Algodex is a highly decentralized marketplace with the orderbook completely on the Algorand blockchain itself. This would make it possible to analize the whole order book data using the Flipside Algorand tables.

All Algorand Standard Assets (ASA) on the Algorand blockchain are supported in Algodex by default.

By querying the mentioned table with the conditions above and the timeframe defined in the bounty, I was not able to obtain any results. At this point I assumed that the data was not completely updated to the table, so I decided to dig deeper on the algorand.transactions table. After some trials I was able to identify the tx_message:txn:apid parameter as the app_id, so I created the Daily tx_id count below to check what happened on the launch. From the activity prior to February 9th, which I assume as Testnet transaction or similar, the significant increase in February 10th and the official announcement on Algodex Blog from February 8th, I will consider the 24 hour period for the analysis as February 10th.

After some discussions on Discord, I managed to find a way to identify the buys and sells by extracting the decoded string info on the tx_message:txn:note parameter. I used some code from user @The Laughing Man to decode this info and look for the escrowOrderType parameter to filter between buys, sells and closed order types.,

Assuming buys as order in which the user exchanges ALGO for any ASA and sells as order in which any ASA is exchanged for ALGO, I will calculate the total buys and sells as filled and unfilled orders for each type. A differentiation between filled and unfilled will be done combining transaction count and order types.

To answer the total amount of ALGO volume, I looked for the TX_MESSAGE:txn:amt parameter for all transactions on February 10th and converted it from microALGO into ALGO, aggregating all results for the analysed day.

Comparing the total transactions for each escrow (ALGO/buys and ASA/sells on February 10th with the order type from below:

• ALGO buys were 2239 (Daily tx_id count for Algodex)
• ASA sells were 3093 (Daily tx_id count for Algodex)
• ALGO closed buys were 290 (2239 - 1949)
• ASA closed sells were 378 (3093 - 2715)
• These two closed amounts add up to 668 (see table below) which were filled
• Total unfilled orders are 4664 (2715 + 1949)

Although ALGO buy orders are less than ASA sell orders, ALGO buys have 10 times more volume than ASA sells. Interestingly, close order type have no amount under TX_MESSAGE:txn:amt, maybe this info is saved under a different parameter for closed orders.